Summarized from UseNet groups:
comp.publish.cdrom.hardware
comp.publish.cdrom.software
comp.publish.cdrom.multimedia
DISCLAIMER
This FAQ is presented "as is" and the views expressed in it are those of the author and are not necessarily endorsed by CDROM2GO
Sections:
I have owned an HP SureStore 4020i CD-R drive for about six months now. During that time, I've monitored, and sometimes participated in several Usenet discussion groups that deal with topics regarding burning CDs. Each day I read several questions that have been asked time and time again, that I know the answer to, but just don't have the time to respond to. The following is a compilation of some of the questions that most often appear on the comp.publish.cdrom.hardware, comp.publish.cdrom.software, and comp.publish.cdrom.multimedia UseNet Newsgroups. Formatting for this FAQ was borrowed from the comp.dcom.lans.ethernet FAQ.
I will try to post this FAQ monthly, or more often when I feel that sufficient information has been added to warrant redistribution.
This FAQ is no longer maintained.
The questions and answers are mostly summarized from the UseNet Newsgroups comp.hardware.publish.cdrom, comp.publish.cdrom.software, and comp.publish.cdrom.multimedia. Specifically the following people have contributed to this FAQ. Their knowledge and experience is gratefully acknowledged.
New contributions, suggestions and corrections should be mailed to the current FAQ maintainer, who is listed in Q&A 01.03 above. However, you should note that if you are submitting a correction you must provide both the old and suggested new text—messages to the effect of "this is wrong, fix it" will be ignored.
You may freely distribute this document for non-commercial purposes as long as the contents remain unchanged (including credits) and you do not gain any direct profits from the distribution.
Right here! :) I, Greg Volk, am not an expert in the field of compact disc technology. Because of this, some or all of the information in this document may be incorrect. Furthermore I will not be held responsible for the use/misuse of the information contained in this document.
I have written many (100+) audio and data CDs with the following hardware/software combination.
While this setup does work quite well for 99% of what I burn, I have run into one underrun problem when I was burning a CD that would have contained 630mb spread over 11,000+ files. The CD did not finish, and I was left with a half written CD, and an underrun error. To remedy this, I dropped down to single speed burning, for this one disc, and it worked just fine. The point I'm trying to make, is that good, fast IDE drives (such as the WD31600A) are good to burn from unless you have a truly massive amount of small files. I still burn stuff at 2x, however when I have more than 7,000 files to burn onto one CD, I drop down to single speed just to be safe.
I imagine that the solution to this would be to go to some sort of SCSI drive, however I do not have the money for this. Another possible solution would be to build an image of all the data that I wished to burn, and then burn it from the image. I lack the extra drive space to do this, so it is not plausible. Yet another solution that might eliminate the 11,000 file underrun problem would be to upgrade my PC to something quicker than a 486-66.
Additionally, Deirdré Straughan had the following to add:
It's not unusual to suffer buffer underruns when writing large numbers of small files, just because it can take more time to locate the file on hard disk and open it than to actually write it. Defragging your hard disk may help in this situation -- at least there are fewer pieces of the files to search for.
A: The answer to these questions, and more regarding hard drives for high performance demand audio/visual/data use can be found in the following article by Bertel Schmitt.
by Bertel Schmitt
used with permission
Hard drives have been blamed for just about any ill in the digital audio & video editing world. If frames drop or audio flakes out, the stock answer of any harried tech support person is "it must be your hard drive."
Users on the other hand cant understand why they have just shelled out $$$ for a super fast HD that supposedly shovels 7 MB/sec, and all they can capture is 1.8 Mb/sec (if they are wearing striped socks and the moon-phase is a zero crossing).
The correct answer is a long one. It involves looking at all phases of the capture chain, at bus & CPU saturation, at PCI chipset designs, at bursting, at system stalls, even at memory access. Audio and video is a steady stream, but PCs are not configured to handle streams well. If anything goes wrong at any point of the capture chain, a data stall occurs and frames drop.
Instead of looking at the big picture, blaming the poor hard drive became en vogue. The biggest bugaboo and whipping post is the dreaded thermal recalibration, tcal for short. If someone goes online and says "I captured for 10 seconds and my frames drop," you can be sure that someone answers: "You are the victim of the heinous thermal recalibration. Go out and buy yourself an AV disk."
Some drive manufacturers quickly capitalized on this. "AV Drives" became the gold plated MonsterCables of the digital world. People happily pay $100 more for the same drive, as long as it has "AV" attached to the part number. In a world of eroding margins, "AV" and "thermal recalibration" became a god-sent to struggling drive manufacturers and system integrators. But do AV Drives really perform wonders as advertised? Have they slain the dreaded tcal? To cite Kris Kristofferson: "It's a walking contradiction, partly truth and partly fiction."
What is thermal recalibration and why do people say such awful things about it? Run your hard drive for a while and touch it. Its hot. A change in temperature leads to expansion or contraction, in a hard drive, it changes the geometry of the platters. Modern hard drives reserve one dedicated servo surface. On startup, the drive reads the servo tracks. After extended usage or after an error condition, the drive recalibrates. Recalibration takes approximately 40ms per surface, depending on the size of your drive and the number of surfaces, a complete recalibration can take somewhere between 0.2 and 1 second. The bigger the drive, the longer the tcal. During the tcal, nothing is being read or written, so unless other precautions are being taken, the data stream is being interrupted. Until recently, improving the performance and capacity of the drive involved adding more platters and raising the rotational speed. Sadly, adding platters or raising the rotational speed means more heat. The main heat generator in a drive is friction between the platters and the air that surrounds them. More platters, more speed, more heat. That, for instance, was the reason why one never saw 9 Gig Seagate Elite or Micropolis 1991 - - for years the mainstay in AV editing circles - - with 7.200 rpm. They rotate at 5.400 rpm, because, as one engineer told me, "if they would spin any faster, they'd probably melt away."
By the same token, the best way to keep your hard drive healthy, wealthy, wise and reasonable free of tcals is to blow a lot of air over it. I have a stack of five Barracuda 4s, one as the system drive, four as a striped set. Uncooled, you could cook with them. A 4" inch fan in front and back keeps them at a moderate temperature. The fans also lengthen their lifespan.
A common myth is that current technology AV drives do not recalibrate. That's baloney, unless you are one of the lucky few who could lay their hands on a drive with embedded servo tracks and magnetoresistive (MR) read heads (MRH). This technology, pioneered by IBM in 1991 (who, as usual, didn't exploit it enough - have you heard much about AV drives by IBM?) is now going mainstream. In the second half of the year, Seagate, Fujitsu, HP et al will ship embedded servo MR drives in quantities, instantly obsoleting this article. MR drives will also increase the data density and hence the data thruput by 60% to 80%. At the same rotational speed of 7.200 rpm, one of these drives will deliver approximately 12 MB/sec instead of the current 6.8 MB/sec.
That's why these drives will ship as wide or ultra SCSI drives. More on the matter in the WWW under
http://www.seagate.com/new/newtop.shtml.
Micropolis claims that their "hybrid servo system combines the best features of both dedicated and embedded servo system designs and completely eliminates the requirement for periodic T-Cal operations." More by pointing your web browser to http://www.microp.com/AVG.html. Micropolis has implemented this technology in their "Gold" series. Note that they are saying: "completely eliminates the requirement for periodic T-Cal operations." They don't say that they have ditched tcal altogether. In a pure embedded servo design, no recalibration is necessary at all. All tracks contain servo information and each head will be brought automatically to the correct position without any interruptions.
As we have seen, current technology AV drives do have to tcal eventually. They just do it differently than regular drives. A regular drive does periodic maintenance, or it goes into tcal when it sees any (usually correctable) error requiring a re-read. AV-drives try to hide the recalibration. They perform maintenance while the drive is idle, they interrupt maintenance when the drive gets busy, or they do what's called a "posted tcal:" If a recalibration is scheduled and the drive is busy, the drive logic delays (or posts) the recalibration until an idle period is reached. This is the most important feature of AV drives. Its a feature that's shared by most modern high performance SCSI drives, AV or not. A stock ST15150, a.k.a. Seagate Barracuda 4, does posted tcal as a factory default, just like its more expensive brethren with the "AV" suffix.
The grizzled backyard mechanic is used to tuning a car. A hard drive can be tuned within limits. Just like a car can (or, until computers took over, could) be tuned for power or economy. Don't expect too much from tuning (10% on a good day). And as with the car, power carries a price.
First, a little background. SCSI drives are better and more powerful than their IDE brethren, because they have a little computer and some memory on the drive itself. The controller card writes data into the memory, issues a command to the on-drive computer and then goes about the rest of its business. The computer on the drive does the rest. The computer on the drive is programmable and it can be directed to change its attitude towards data. That's done by setting variables or flags, which are stored in so-called "Mode Pages."
One of the best known mode page setting is the Write Cache Flag. Most SCSI drives have an on-board cache which ranges from 256K up to 1 Meg and more. On some drives, this cache is strictly for reading data, on most newer drives, the cache can buffer reads and writes. Obviously, in a capture situation, a cached write is better, because the computer simply writes into the cache and doesn't have to wait for the drive to complete the write operation. Most drives are shipped with the write cache disabled. That's because in normal life, more than 90% of all drive operations are reads and therefore the whole cache is dedicated to reads. Some drive manufacturers will also claim that disabling the write cache is safer in case of a power failure, but that's a Red Herring. A power failure during a file write is a ticket to disaster, itty-bitty write cache or not.
There are several utilities floating around that allow the enabling of the write cache. One of the best and painless is the SCSI Explorer by Adaptec, which is part of their EZ-SCSI package. Free with their controllers.
The rest of the AV tuning has been done by the hard drive manufacturers themselves or by "Speed Shops", a.k.a. hard drive integrators.
Their basic strategies are:
As you can see, half of the "secret tuning recipe" sacrifices speed for data integrity. The theory behind it is that a flipped bit in a video stream will have much less impact on your well-being than a flipped bit in accounting data. That's why "hot" AV drives should only be used for AV, not as a system drive. But even as an AV drive, fiddling with error correction can be an invitation to disaster. If a bit is flipped in video or audio data, you will hardly hear or see it. But if its flipped in a file pointer, the whole file could be burnt toast.
The tuning of AV drives used to be a specialized chore for professionals. With reason, because one wrong setting in a mode page can send the whole drive to the scrap pile. Peripheral Test Instruments in Lakewood, CO, has put out Dr. SCSI, a fairly safe application (unless you perform unguided brain surgery in your drives mode pages - which you can), that performs AV tuning with a mouse click. The program is available for DOS, Windows and NT and it costs less than the premium of one AV drive alone. Information can be obtained via 303-763-7488 info@scsitools.com or http://scsitools.com .
The manual comes complete with in-depth SCSI theory and ample warnings for the unwashed. Highly recommended.
A write emulation test is the best method. When the writing light on the 4020i blinks as it reads all of the files that will be burnt. Unfortunately the current Easy CD for Windows 3.11 that ships with the 4020i does not support the write emulation test. The only test it has is a Speed Test, which produces largely irrelevant information, and can cost you in underruns when you do actually begin burning. The so called Speed Test that you can perform in the Win 3.11 software simply copies all of the files that you wish to burn to null and then sums the total amount of data and the total amount of time, divides one into the other and gives you an average speed in kilobytes per second that the files were transferred from your hard drive to a null device. For the most part, this is a worthless number. I say this because while the average transfer rate of your hard drive may be well above the 300k/sec rate required by 2x burning, what about the lowest rate? It is with the lowest transfer rate that we run into difficulty. Who cares about the average? If the speed test would tell us what the lowest rate was, then we could decide whether or not to burn at 2x, 1x, or not at all.
Rick Adams contributed the following regarding drive temperature:
Besides the fan built into the computer's power supply inside the case, I've got a fan inside that moves more air through the case. My case provided a place to mount this, so I just mounted one.
The Fujitsu hard drive M2934QAU I have gets hot and the fan I've added helps this.
Note that the fan on the CPU chip doesn't cool the case. It just keeps the CPU chip cooler by moving the heat away from the CPU, avoiding a hot spot on the chip.
In HP's specs listed in the 4020 installation guide for the CD-writer drive, the maximum operating temperature is shown as 35 degrees C, which is 95 degrees F. This means that if the internal temperature inside your computer's case goes above 95 degrees F, proper operation of the 4020i is not guaranteed. It doesn't mean it will always just stop working. It simply means that HP doesn't suggest operation above that temperature.
I measured the difference in temperature between the inside of my computer's case, and the room, with and without the extra fan I installed turned on. With the extra fan turned off, using just the fan built into the power supply, the difference was 22 degrees F. With the fan on, the difference was 8 deg F.
This means that to not exceed HP's operating temperature of 95 I had to keep the room below 73 deg F without my extra fan, or 87 deg F with my extra fan. Quite a difference.
I haven't tried this experiment with a cooler running hard drive, but I know that other large fast scsi's get hot too. If you've got one of these, there's a chance your internal case temperature is exceeding HP's maximum for the CD-writer unless you've got a supplemental fan.
Comparing this to other equipment, it very well could be that the 4020i's 95 degree requirement is stricter than any other devices in your computer. I note that most ICs (like RAM) for consumer application are rated to 40 deg C, or 104 deg F. My Plextor 8X CD reader is rated to 45 deg C or 113 deg F, for example.
Radio Shack sells a thermometer module 277-123 for about $20, but a fan only costs $10 or $15 bucks, so you might save the expense of measuring by just getting the auxiliary fan which is cheaper! Make sure you've got a place to install and connect it!
As far as I know, the 4020i was first shipped with version 1.14 firmware. From my experience, this original firmware, as well as the original release software, was buggy and was quickly replaced with version 1.20. I downloaded the 1.20 flash upgrade, and attempted to run it. According to the dialog boxes, it claimed to have completed successfully. However when I tried to burn CDs with the drive, I encountered several error messages, and when CDs did burn, they usually wouldn't close correctly. I called HP about this, and they sent me a new drive, with 1.20 firmware on it.
Version 1.20 firmware seems to be quite solid. I have burnt over 100 CDs with 4020i drives using this firmware and haven't had any problems. Thus when the 1.25 flash upgrade appeared, I was rather reluctant to upgrade. However I'm told that version 1.20 does not do CD-XA discs correctly, and that 1.25 was the answer to this. Hence my reason to upgrade. I ran the flash upgrade, and it claimed to have flashed just fine. Unfortunately when I would try to execute a CD-Info command on the drive, error strings in error dialog boxes would appear. Per HP's tech support instructions I re-downloaded the 1.25 flash upgrade, and flashed the drive again. Once again, it completed successfully. This time however I was able to use the drive. The CD-Info command worked as it should and everything seemed to be ok. When I tried to burn a CD, errors once again became abundant. CDs would not close correctly, usually soft locking Windows 3.11. I called HP once again, and the support rep told me to download the 1.20 firmware and flash back to it. I did this, and was able to write several CDs successfully after doing so. However, about five CDs later I began getting the error "Calibration Area Full" from the drive, just before it would begin writing. I called HP about this, and after questioning me about what brand of CDs I was using (this error occurred on every blank I tried, HP, Kodak, MEI, Verbatim) the tech decided to send me a new drive. Oddly enough, the new drive has version 1.20 firmware on it, not 1.25.
To summarize:
Read on to see what version 1.25 and 1.27 do.
Jim Watson asked HP what version 1.25 changed, and received the following reply:
Firmware 1.25 improves overall compatibility and enhances performance. It also allows disc at once recording and improves access to video CDs and CD+ discs.
Any problems with the drive that were advertised by users like Jeff Arnold have been fixed with 1.25.
Toni Lindroos asked HP what version 1.27 changed, and received the following reply:
Firmware 1.27 improves overall compatability (even in Windows NT) and enhances performance. It also allows disc at once recording and improves access to video CDs and CD+ discs.
Please note that you MUST upgrade 1.14 drives to 1.20 before writing 1.27, and you MUST use our SCSI controller in Windows 3.x or 95 to do the upgrade.
Jim Watson posted the following on the comp.publish.cdrom.hardware UseNet news group.
I emailed the tech support at HP and apparently this upgrade is only for certain situations text follows:
Q. Can you please explain why you have password protected the firmware files?
A. Not all users need the 1.25 firmware upgrade. The firmware upgrade programs have been password protected because we need to know who is using them and give special instructions. There will probably not be any more than one more upgrade program in the future. I can offer passwords and filename information via Email.
— HP SureStore Tech Support.
Jose A. Cid had the following to say about his difficulty with running flash upgrades:
I would recommend everyone use an active terminator. I could not flash a ROM upgrade until I changed the termination per advice from an HP tech support rep.
It has been my experience that Target Abort errors are generated by the Advansys SCSI card that the 4020i ships with. To successfully use jarnold's FILE2CD software, use a different controller. After getting my first Target Abort error, I decided to attempt to use an Adaptec 1542 with the drive. This appeared to completely remedy the problem. I have used the FILE2CD program and 1542 combination many times since then and have had nothing but successful results.
However, Darin Johnson had the following to say about his NCR SCSI adapter, DAO (commercial FILE2CD), and the 4020i:
I have recently purchased Mr. Arnold's DAO program and it has worked for me except that I started getting those -24 errors (Target Abort). So I took his advice and took out that Advansys card, and strung the 4020i onto the PCI NCR SCSI-2 card that also accommodates my hard drive, 2nd CD-ROM drives, and an 88Meg Syquest drive. This seemed to work better, however, I'm still getting Target Abort errors.
Having read that, it is clear the Advansys card is not entirely at fault, and that the problem only occurs with certain combinations of the 4020i and various SCSI boards.
Furthermore, Russell Thamm had the following to say regarding the 1542 and Target Abort errors:
As far as I can tell, the target abort error returned by Jeff Arnold's program could either be:
I also had problems with EasyCD with write append errors, so I presumed that the problem with FILE2CD were also 'write append' errors.
The point is that I had these problems using an Adaptec AHA-1542 interface card. When I had the HP4020i replaced AND upgrade to the 1.25 firmware, the problem has disappeared (I hope). As far as I can tell, the fault was caused by either a dud unit or the 1.20 firmware.
Yes you can. I have successfully used the 4020i with an Adaptec 1542CF controller. The only change that I had to make was in the SCSISelect utility. Upon getting into SCSI select, you will need to select Device Configuration. Then go to the ID that the 4020i is on (default = 2). Where it says YES for FAST SCSI, change it to say NO. If you attempt to burn CDs when this is set to YES, you will run into mysterious Buffer Underrun Errors. I don't know why this causes buffer underruns, but it does.
As for using SCSI controllers other than the 1542CF, I'm looking for others who have done this. If you wouldn't mind typing up your experience with other SCSI controllers and sending them to me, I'd be happy to add it to this document, and give credit. Tim Goldstein sent in two things to watch for when purchasing a host adapter for your 4020i drive:
Rick Adams had the following to say about the 2940UW and 4020i combo:
I have an Adaptec 2940 UW ultra wide, which isn't the same as a 2940 regular. Using the 1.25 firmware I was able to get the HP drive to read correctly when connected to the 2940UW, but I had many CD writing lockups and errors so I now use the HP supplied controller and 1.20 until the HP / Adaptec ultrawide / firmware / whatever / issues are resolved!
Rick also points out a configuration problem with the 2940 that can cause buffer underruns.
I did get a buffer underrun once, but since I always write CD's from a freshly formatted partition on my hard drive, I had to dig further to discover that Adaptec's default 20 MB/sec ultra wide transfer rate was pushing my Fujitsu 2934QAW drive too fast and causing SCSI bus error retries. The SCSI bus retry procedure is slow enough to cause the buffer to underrun. Setting the maximum synchronous transfer rate that the 2940 uses to talk to my drive down to 10 MB/sec cured that.
Yes, you can have other devices on the chain, providing you make sure that your termination is correct. I have run my 4020i off of an Adaptec 1542CF with a scanner plugged into the external SCSI port. The scanner was terminated as well as the 4020i. This setup appeared to work just fine.
Tim Goldstein sent in the following about other devices on the same chain as the 4020i:
You must be careful when placing the CD-Writer on a SCSI bus with other devices. If there is a possibility that one of the other devices will be accessed during a write, (i.e. a hard drive) it could cause a data delay resulting in a buffer underrun.
In response to this question, Andy McFadden writes:
Possible resolutions to this problem that occurred while using an Adaptec SCSI controller are:
Jose A. Cid had the following to say regarding termination:
I would recommend everyone use an active terminator. I could not flash a ROM upgrade, until I changed the termination per advice from an HP tech support rep.
Rick Adams posted the following to Usenet about the Advansys SCSI host adapter:
If you are using the HP supplied controller from Advansys be advised there are at least 2 revisions of the card. The one I first got was positively identified to produce random read errors when reading from the HP CD. This means that if I used the HP drive to read a CD, I occasionally had corrupted data during reading. Didn't matter if the disk was a silver manufactured CD or a CD writer burnt CD. I was sometimes seeing as many as 10 bytes wrong in one whole disk read, to as few as 1 byte wrong in 30 CD reads. Yes this is a tiny error rate, but there should be absolutely no errors of this sort. This was traced down to be the cause of one CD I burnt that had two wrong bytes on it. That is, since the drive read in some wrong bytes, that's what it wound up writing.
The solution to the read error problem was to swap out the HP supplied controller card. The one that caused the read errors had the following chips: two of IS61C256AH-20J, two of ML6509CD, one of Atmel 16FE-17, one of Advansys ASC900, one of IS93C46-3GR. This board was sent back to HP.
The easiest way to identify if you have the better board which HP sent me is to see if it has LS245 chips on it. If so, you're in luck. That board didn't produce the above mentioned readback errors.
J. Robert Sims, III contributed the following:
The errors described only happen with particular boards that improperly expect extra lines be driven by the card. Either revision of the board with most motherboards do not have the problem; only the specific combination of the old board with a very small set of motherboards will show the error.
Originally the 4020i shipped with beta software of Adaptec Software Products Group'sEasy CD for Windows v3.11. This original beta version was just that, a beta. It appeared to have several bugs that effected disc closing, and data preprocessing (what it does with all the filenames before it begins to write).
The current version of Easy CD that can be found on HP's web site seems to be solid. I have burnt 100+ CDs with this version, and most have been successful. Of those CDs that did not burn correctly, I have concluded that the source of the problem was hardware related.
Because Sony doesn't want you to. Sony has changed the format several times in the last few months, to make it more difficult for CDR software packages to create successful images of these discs. Several packages will make copies of older Play Station games, but the new PS discs have been written/stamped differently to make this impossible without modifications to existing CDR software packages.
Deirdré Straughan wrote the following in response to a similar question posted to comp.publish.cdrom.hardware.
We disabled 4x reading on the HP/Philips 2000 because in testing we found that an "Optimum power calibration failure" error occurs when, after reading at 4x, you write at 1x. It seems to be a firmware problem. we have informed the manufacturers about it.
Jeff Arnold was a developer of CDRWIN, a CD-R burning program released by GoldenHawk Technology. The software was last updated in 2008 and appears to be abandoned as of July 2008.
…before it begins writing, I have gotten something similar to the following error:
Virtual database error 39-00-00-00
What is going on, and what can I do about it?
According to the Easy CD error list, this error means that the maximum amount of data that can be put on the CD has been exceeded. Remove a few files and try it again. The Max CD size is not dependent on the number of files.
This error is possibly linked to a humming noise that the drive makes after heavy long term usage. I called HP Tech Support regarding this, and for a while it seemed that they were attempting to blame the problem on the CD-R media I was using. After getting the same results with the HP SureStore CD-R brand media they agreed to send out a new drive. Unfortunately this is the extent of what I know about this error message.
Deirdré Straughan posted the following article to Usenet explaining buffer underruns, and what can be done about them:
CD writing is a real-time process which must run constantly at the selected recording speed, without interruptions. The CD recorder's buffer is constantly filled with a reserve of data waiting to be written, so that small slowdowns or interruptions in the flow of data from the computer do not interrupt writing
A buffer underrun error means that for some reason the flow of data from hard disk to CD recorder was interrupted long enough for the CD recorder's buffer to be emptied, and writing was halted. If this occurs during an actual write operation rather than a test, your recordable disc may be ruined
Hard Disk
Hardware
Memory Resident Programs
Networks
Files to Be Recorded
Other
Virtual Memory (if you have more than 16 MB of RAM)
Hard Drive Typical Role (if you have more than 16 MB of RAM)
Turning Off Auto Insert Notification
Note: You should do this for every CD unit on your SCSI bus, including the CD recorder itself!
The above contribution from Deirdré Straughan is © 1996 Adaptec, Inc., all rights reserved.
Please visit the manufacturer's website for a list of fixes to common Easy CD problems.
…the 4020i stops and says "write append error." Also, when I try to select "Close Disc" the drive starts writing but then stops and reports "Internal Controller Error." What going on here?
Sa-Nguan Uarsirisab sent this question and the following answer which was given to him by Hewlett Packard SureStore tech support:
Write Append and Internal Controller Errors are usually caused by bad SCSI card settings, bad cables, and/or memory issues.
Use a different SCSI cable if one is available. Otherwise switch ends of the current calbe so that the connectors are hooked to the other device. Check for third-party memory managers like QEMM or RamDoubler. These can cause problems Windows 3.x and should not be used in Windows 95.
Rick Adams mentioned the following regarding Internal Controller errors:
Just prior to writing a CD, I power down and restart my computer into Windows 95. Naturally I make sure screen savers are off and my 10B2 network card is connected only to a 50 ohm terminator. But what really seems to have gotten rid of the "internal controller error" problem was the procedure of first turning the computer off and on, then not doing any other CD reading or starting any other tasks before writing the CD. It seems something isn't being properly reset unless the power is turned off.
Another possible source of "Internal Controller" errors is heat. See 02.04Q for more information about max operating temperatures.
…close the disc and why can I only access the last written session but no sessions prior to that?
Sa-Nguan Uarsirisab sent this question and the following answer which was given to him by Hewlett Packard SureStore tech support:
The software will close the disc if it can. Since it hasn't been able to so far, that indicates communication with the CD-R might not be very good.
Use the CFGISA program to change the controller's DMA Speed to 3. Also try changing the DMA channel and IRQ. Frequently when a session fails, it will disappear when you add more data to the CD later.
I (gvolk) have also found that sometimes Jeff Arnold's freeware program FINALIZE can help in correctly closing an errored CD.
Sa-Nguan Uarsirisab sent this question and the following answer which was given to him by Hewlett Packard SureStore tech support:
Some computers will attempt to shut down hardware to save power. This action has been seen to cause delays in writing, even when power saving features are suspended through software. The best fix is to disable power saving features in the computer's CMOS (BIOS) setup.
…only the most recent session shows up, but in other CD-ROM drives all the sessions are visible.
J. Robert Sims III answered this one with the following:
The c4324hlp.vxd driver is not installed in the \windows\system\iosubsys directory. Get the latest version of the driver. The driver cdr4vsd.vxd from Adaptec is a newer substitute for c4324hlp.vxd.
Several users have written to me regarding this problem, and before sending back my second drive it began happening to me. For me, it began happening only after I had the drive for a while, (3 months) and after I had burnt/read 100+ CDs. As far as I know, something is obviously wearing out and you should call Tech Support and ask them what they suggest you do. Additionally, see the question dealing with "Calibration Area Full" errors in the SOFTWARE ISSUES section of this document for a possible link to an error message.
Success in burning CDs from a network drive is dependent on many things. Most PC networks today are 10 Megabit per second Ethernet. This means that the maximum rate at which data will get from the host computer to the target machine is theoretically as high as about one megabyte per second. Of course, predicted values and actual values are not always the same. Thus is the case with 10 megabit per second Ethernet. With the right combination of hardware and software, we can typically achieve transfer rates of up to 800 kilobytes per second. This is on an Ethernet segment that has little or no traffic, and when both the host and destination machines are doing nothing but concentrating on the job at hand. Network data flow interruptions can come from various sources on an Ethernet network. Most delays are a result of network traffic. An Ethernet network is not a good choice for CD burning for one major reason. Ethernet is a non-deterministic (non-real-time) "delay or drop" network. This means that when you request data from a file server on an Ethernet network, it may or may not get to you, and there is no way to tell when, or how fast it will get to you. I have burnt CDs on a 10 mb/sec Ethernet network and have had mixed results. Originally I did a music CD (very few, very large files). This worked fine at 2x. The host computer was a P-90 with 16 megs, running Windows 95 reading data that was stored on a 1.2 GB Western Digital hard drive. When I tried to do a data CD that contained 600 megs across 10,000+ files I encountered nothing but underruns. Underruns were prevalent at both 1x and 2x. And this was on a network that had zero traffic aside from the host PC and my PC. Because of this, I am inclined to discourage people from burning CDs from a network drive. I haven't had the chance to work with 100 Megabit per second FAST Ethernet, so I don't know how it performs with this type of task, but with as cheap as large hard drives are, it seems that going the hard drive route ends up being the less costly (in terms of blank CDs, and stress :).
HP's 4020i support page can be found at: HP SureStore 4020i Resources
Drivers, software, and firmware upgrades are available from:ftp://ftp.hp.com/pub/information_storage/surestore/cd-writer/
More information about Roxio Easy CD Creator recording software can be found at the following web site: http://www.roxio.com/
Andy McFadden writes:
You can, but it's not clear that you'd want to. The seek times tend to be slower than a standard CDROM drive because the head assembly is heavier. There's also not much need for rapid seeks when writing a disc, so there's little reason for manufacturers to try to optimize this.
Some users have reported jerky video playback on a CD-R drive.
The MTBF on CD-R units tends to be low, so it may be wise to use a different drive for general use.
Deirdré Straughan writes:
It means that the recorder is unable to calibrate laser power for the disc. Try different media. If that doesn't help, get the recorder checked.
This is a firmware problem, and has been corrected in firmware 1.25. See Q&A 03.01 for more info regarding this bug.
As far as I can theorize, it is because the 4020i drive was one of the first low priced, mass marketed CD-R drives out. The 4020i beat most of the competition to the market for one reason. It wasn't finished. It wasn't ready to be used by the general public. When first introduced, it had buggy firmware, and the original units were shipped with beta software. This is a bad combination for any device, but is particularly bad when your dealing with a device that uses write once media, that costs $8.00 per disc (average). In addition to being reasonably priced, HP's good reputation for quality computer peripherals lead many buyers to purchase a 4020i drive.
Since the drive's original release, HP has released new (non-beta) firmware, and Adaptec Software Products Group has released new (non-beta) software. This has certainly had the effect of silencing many of the critics of the 4020i, but many still remain. Many of these critics are having trouble because they do not have the necessary hardware that is required to write a CD, while others correctly cite existing software/firmware bugs. Having said that, I do believe that the HP SureStore 4020i CDR drive is equally as good as other CDR drives in its price range.
In response to this question, Deirdré Straughan makes the following point:
One reason for this is simply the numbers. Many, many units of the HP are being sold. In my experience, a certain percentage of users of ANY CD recorder and software will have problems, especially in the beginning when they're getting their systems tuned to write CDs. That's real life with computers, folks, and I think you're all experienced enough to know that.
Real life on the Internet is that most people come to forums like this one to seek help and/or complain. Not many bother to pop in to say "I just love this recorder!" So you will always see a large number of people beating up the HP (and other recorders) online. But please keep in mind that they are only a percentage of the people who are actually using the recorder -- the contented majority is largely invisible.
…The source WAV/BIN/MOT files sound perfectly fine when played off of the hard drive. What gives?
In response to a similar question posted on comp.publish.cdrom.hardware, Jeff Arnold replied with:
Just to set the record straight... this person is using a pirated copy of my software which is why he's getting "crackles" in his audio tracks.
I got tired of people pirating my software, so when the program detects that it has been hacked, it randomly writes zeros into the data being written to the blank disc.
The pirates (a group known as X-FORCE) that hacked the protection on my software aren't too bright. They didn't even bother to test the software before distributing it around the net.
I finally got my revenge :-) :-)
…and plan to make more. What can I do to speed up the process of copying the audio CD tracks to WAV files?
Most of the typical CD Rom drives out there can not do digital audio extraction at more than 1x. Attempting to extract audio tracks at more than 1x with these drives usually causes jittering to occur, which ends up giving you an incorrect reproduction of the audio data in your WAV file. Jitter can be blamed for inconsistent extracted file sizes, and popping/crackling noises in the output WAV files. At the current time, the only CDRom drives that I have seen that will do digital audio extraction have been SCSI. I don't know if any IDE/EIDE CDRom drives will do it, or if software exists to do so via these drives.
To get around the Jitter problem, you can do one of two things:
If you are wanting to use your IDE/EIDE CD-ROM drive for Digital Audio Extraction, check out the following web page: http://www.tardis.ed.ac.uk/~psyche/pc/cdrom/CDDA.html
The following is a list of items that you may want to run through before burning a CD as they may have an impact on CD burning.
As of this version of the FAQ, I do not have any substantial information regarding solutions to reading problems. I know that the problems do exist, and are most likely related to shortcomings in software (drivers), but beyond that I have little to offer in response to the above question. If anyone out there has something more informative, please pass it along to me. The problems involving reading are typically described as follows:
Also, see 05.12Q about updated .VXD drivers.
…Everything works great, except when I try to do a CD to CD copy with Easy CD Pro 95. When I have done this, with a Mixed Mode CD (1 Data track, several audio tracks) the resultant audio has various pops and clicks throughout the tracks, but the data is just fine. What's going on?
I (gvolk) don't know. Several users have posted this question to Usenet, as well as asked me about it, but as I am not familiar with Easy CD Pro 95, I am unaware of a solution. If anyone has an answer to this, pass it along, and I will add it to the FAQ.
Version Number | Date | Main reason(s) for new version. |
1.0 | 5/31/96 | Initial distribution. |
1.1 | 6/2/96 | Applied standard Internet FAQ format. Added information regarding AV hard disk drives, and Ethernet as a data source. |
1.2 | 6/6/96 | Added contributions regarding firmware, swapping of left and right audio channels, possible solutions to buffer underruns. |
1.3 | 6/17/96 | Added info regarding Digital Audio Extraction, solution to popping noises when using the Goldenhawk DAO software, problems related to Advanced Power Management (APM), closing discs, and "Internal Controller Error" message. Added URL of Easy CD Error Codes. |
1.4 | 7/1/96 | Added Rick Adams' contribs regarding 2940 & 4020i combo, and possible "Internal Controller Error" remedy. Added URL of software for DAE from IDE/EIDE CDRom drives. Added info about heat related problems, added Pre-Burn checklist, corrected Easy CD Error 39-00-00-00 answer, added URL of "Common Easy CD Errors." |
1.5 | 9/1/96 | Revised answer to 05.07Q. Added 05.12 Added 04.07 Added 06.11 Added 06.12 Revised answer to 06.05 |
1.6 | 10/1/96 | Added Adaptec copyright notice to contribs section. Updated 04.06 Deleted 04.07 Updated 05.12 Added 05.13 |
1.7 | 11/1/96 | Revised 03.01 Answer. Maintainer's (gvolk's) address has changed |