VOGONS


Reply 20 of 53, by BenWit

User metadata
Rank Newbie
Rank
Newbie

HOLY CRAP
Now that is a discovery and thanks for it.
My unit has the same riser card (also with solider in TC4), so this is definitely going to be added high on my retrocomputing projects list. As well as a case mod to make it easier to change out the SD card boot drive.

Reply 22 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++

They are caps.

I just matched the physical size up. I have no idea what UF/V they were.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 23 of 53, by WyndC

User metadata
Rank Newbie
Rank
Newbie
cyclone3d wrote:

They are caps.

I just matched the physical size up. I have no idea what UF/V they were.

I can't PM yet, but Cyclone3d when I populated my board last night, the computer wouldn't boot at all with or without the sound card. Any chance I could pay you to populate mine? labor/materials/shipping I'm based out of MI.

Thanks.

Reply 24 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++

Yeah, I guess we could work something out.

Go ahead and post hi-res pics of the front and back so I can see if I see anything wrong first.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 26 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++
franky52 wrote on 2019-12-07, 23:34:

I’m suffering from very low sound and with a lot of noise.

Does this fix solve both problems?

Yes, it fixes both problems.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 27 of 53, by franky52

User metadata
Rank Newbie
Rank
Newbie
cyclone3d wrote on 2020-01-23, 03:20:
franky52 wrote on 2019-12-07, 23:34:

I’m suffering from very low sound and with a lot of noise.

Does this fix solve both problems?

Yes, it fixes both problems.

Thank you. I will try to populate it.

Reply 28 of 53, by xjas

User metadata
Rank l33t
Rank
l33t
cyclone3d wrote on 2018-12-02, 08:01:

I have figured out what the problem is with the AW744-L2 card and the HP T5720 thin client.

It is PCI bus noise.

Fully populating the capacitor places on the PCI riser card completely fixes the issue. I am guessing this will also fix issues with "compatibility" with other cards as well. It should also significantly clear up the audio output if you are using a different sound card.

So, I'm in the same boat here with my Fujitsu Futro S300. Neither my "Creative" CT4810 (AudioPCI) nor my AW840 / CMI8738/PCI-SX want to work with it (both cards are known good - actually the Creative DOS drivers init the card and upload the wavetable sample ROM sucessfully, but no sound output is to be heard. The CMI card just plain won't initialize.)

Fortunately the solder pads & holes are all nice and clean, but unfortunately, there's no published values for the caps anywhere. This is the riser:

riser.jpg
Filename
riser.jpg
File size
278.78 KiB
Views
1387 views
File license
Public domain

The part # on the riser is 736TR3230K100, incidentally, not that that helps much.

How can I determine the values of the needed caps? Anyone want to take an educated guess?

twitch.tv/oldskooljay - playing the obscure, forgotten & weird - most Tuesdays & Thursdays @ 6:30 PM PDT. Bonus streams elsewhen!

Reply 29 of 53, by zolli

User metadata
Rank Newbie
Rank
Newbie

I'm struggling with this, also. I have two T5720 machines with the PCI riser + expansion cases. I would really like to populate the risers to spec, if possible. I took a look at the PCI Local Bus Specification version 2.2 (That's the version I found through Google.) and in section 4.4.2.1 I found (apologies for the screenshot; the PDF uses embedded fonts):

PCI_spec.png
Filename
PCI_spec.png
File size
88.42 KiB
Views
1309 views
File license
Fair use/fair dealing exception

A few questions:

  • Is there a way to determine what PCI version the T5720 adheres to? Does the T5720 support 3.3V, 5V and universal PCI cards?
  • Are the intended capacitors found on this PCI riser meant to deal with the decoupling of expansion boards or are these used for some other reason? (I would have assumed that the decoupling capacitors would be found on the PCI expansion board itself, not on this riser.)
  • Is there any indication that the PCI boards that work vs. the PCI boards that don't work has something to do with being a 5V, 3.3V or "Universal" expansion board?
  • If these capacitors are meant for decoupling, the PCI spec (table 4.13) suggests that there should be as many as 27 capacitors (pins 5A, 5B, 6B, 8A, 10A, 16A, 19B, 21A, 25B, 27A, 31B, 33A, 36B, 39A, 41B, 43B, 45A, 53A, 54B, 59A, 59B, 61A, 61B, 66A, 70B, 75A & 79B) but there are only 10 (if I counted correctly) capacitor solder points on the PCI riser. Am I reading the spec wrong?

Reply 30 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++

The capacitor values are not critical. Only thing I would worry about is making sure the rated voltage is equal to or higher than the PCI voltage.

I already posted what I used when I found the fix... at least for the electrolytic capacitors. The SMD that I used, I just matched up the size that would fit.

For all I know, a single capacitor would fix the problem but I just populated the cap positions on the riser.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 31 of 53, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++

Are any of the capacitor pads connected to data pins on the PCI slot? Termination impedance will generally be more critical than power plane decoupling.

All hail the Great Capacitor Brand Finder

Reply 32 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++

I'll have to double check, but I don't think so.. at least not for the electrolytic capacitors. That is how I determined where to use 6.3v and 16v caps.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 33 of 53, by zolli

User metadata
Rank Newbie
Rank
Newbie
cyclone3d wrote on 2020-08-27, 02:12:

I'll have to double check, but I don't think so.. at least not for the electrolytic capacitors. That is how I determined where to use 6.3v and 16v caps.

Thank you for the continued help. Can you explain the above statement: How did you determine where to use 6.3V & 16V caps?

Reply 34 of 53, by zolli

User metadata
Rank Newbie
Rank
Newbie

I tried this mod. I used 47 pF surface mount capacitors and used capacitance and voltage ratings for the electrolytic caps as described in this thread.

I tested the board with a friend who is an Electronics Engineer. With multimeter in hand, we actually de-soldered and re-soldered a few caps thinking that we had a solder bridge. (I'm ok at soldering but not the best.) After everything looked good, I put the riser into a known working system with a known working PCI graphics card.

The result:

burned_motherboard.jpg
Filename
burned_motherboard.jpg
File size
1.31 MiB
Views
1170 views
File comment
Fried T5720 Thin Client
File license
Public domain

The board caught fire. The flames were probably about 2-3 cm high. High enough that the flames scorched the parallel port connector, as you can see in the picture. I was able to blow out the fire before anything worse happened. I shouldn't have powered this device on while it was laying on a carpeted floor! Luckily, there was no further damage beyond the machine.

I don't know what the cause of the fire was. Likely, it was a faulty job on my part but given the oversight of an EE, I'm not sure. Just wanted to post this here to warn others to be careful. If you do this mod, at least consider powering the system on somewhere safer, like a garage.

Reply 35 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++
zolli wrote on 2020-09-01, 17:27:
cyclone3d wrote on 2020-08-27, 02:12:

I'll have to double check, but I don't think so.. at least not for the electrolytic capacitors. That is how I determined where to use 6.3v and 16v caps.

Thank you for the continued help. Can you explain the above statement: How did you determine where to use 6.3V & 16V caps?

By the PCI slot pin voltage.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 36 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++
zolli wrote on 2020-11-03, 18:43:
I tried this mod. I used 47 pF surface mount capacitors and used capacitance and voltage ratings for the electrolytic caps as de […]
Show full quote

I tried this mod. I used 47 pF surface mount capacitors and used capacitance and voltage ratings for the electrolytic caps as described in this thread.

I tested the board with a friend who is an Electronics Engineer. With multimeter in hand, we actually de-soldered and re-soldered a few caps thinking that we had a solder bridge. (I'm ok at soldering but not the best.) After everything looked good, I put the riser into a known working system with a known working PCI graphics card.

The result:
burned_motherboard.jpg

The board caught fire. The flames were probably about 2-3 cm high. High enough that the flames scorched the parallel port connector, as you can see in the picture. I was able to blow out the fire before anything worse happened. I shouldn't have powered this device on while it was laying on a carpeted floor! Luckily, there was no further damage beyond the machine.

I don't know what the cause of the fire was. Likely, it was a faulty job on my part but given the oversight of an EE, I'm not sure. Just wanted to post this here to warn others to be careful. If you do this mod, at least consider powering the system on somewhere safer, like a garage.

Did you try to use this riser in a different system than it came with?

From the one pic, it also looks to me like there was a bulging capacitor on the motherboard.

Did the PCI card work in that same computer before you tried it with the riser?

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 37 of 53, by zolli

User metadata
Rank Newbie
Rank
Newbie
cyclone3d wrote on 2020-11-07, 06:23:

Did you try to use this riser in a different system than it came with?

No. The riser was from the same system.

cyclone3d wrote on 2020-11-07, 06:23:

From the one pic, it also looks to me like there was a bulging capacitor on the motherboard.

That may be a trick of how the light is coming off the top of the capacitor. None of the capacitors on the board are visibly bulging.

cyclone3d wrote on 2020-11-07, 06:23:

Did the PCI card work in that same computer before you tried it with the riser?

Yes. I tested with a graphics card & everything worked as expected.

Reply 38 of 53, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++

Did you have the polarity of the caps correct? Were the caps tested beforehand to make sure they weren't shorted?

Are you 100% sure that the SMDs were caps and not something else such as resistors?

Pics of the riser would be nice.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 39 of 53, by LightStruk

User metadata
Rank Member
Rank
Member

I am having a similar problem with all of my PCI sound cards in my Neoware CA10 thin client, including an AOpen AW744L II. What I don't know is if the issue is simply PCI bus noise or if it's the lack of -12V on the PCI slot. More details in the thread.

Also, my riser card has more than just unpopulated caps - there's a spot for an IC, an inductor, and some resistors too. I have no idea what to put there.