16-Bit I/O Buffer Wing Issue

[madscifi]

I just tested a 16-Bit I/O Buffer Wing that I purchased from LinkSprite and it does not appear to limit the pass-through voltage to 3.3 volts.

Test procedure (number in parentheses is the actual meter readings):

3.3 (3.297) volts/ground connected to the 3.3 volt and ground power connectors.

A volt meter was connected from pin A16 (pin 26 on the chip) to ground.

3.3 (3.297) volts was applied to JP7 pin 8 (pin 23 on the chip, labeled as 15 on the board). 3.3 (3.297) volts was shown on the volt meter.

5 (4.73) volts was applied to JP7 pin 8 (pin 23 on the chip, labeled as 15 on the board). 5 (4.73) volts was shown on the volt meter.

Is there something I'm doing wrong?

(Full disclosure - I've exchanged emails with Gadget Factory in the past about this board because I had concerns about the chip used on this board, but the board was out of stock at the time so I was unable to test an actual board at that time.)

[Jack Gassett]

I'm seeing the same thing here... I know this was working before, I did thorough testing before I manufactured the first batch of these boards. But sure enough, it's passing 5V through here too. The only explanation I have is that I was using an IDT chip before, maybe that chip worked and this SN74CB3Q16245DGGR chip is different enough not to.

Level translation is working with the 5V SN74CBT16245DGGR on other prototype boards I have and the CBT chips are advertised as level translator chips. I redesigned the Buffer Wing to use the CBT 5V part and a Diode to limit the voltage to 3.3V.

The updated design is in Github, I've already ordered a batch of boards to verify that it works as expected, and I will try to get a new batch of boards manufactured as soon as possible.

In the meantime I took the Buffer Wing out of the store and if you have a I/O Buffer Wing please stop using it for now. When the new batch of boards are ready I will replace your old board with one that works correctly. Please contact us at support@gadgetfactory.net with any concerns. I'm very sorry for this issue and will do my best to make it right.

Sincerely,

Jack.

[alex]

Looking at the datasheets you linked to, both the SN74CBT16245DGGR and SN74CB3Q16245DGGR are so called FET switches where each line between the A and B ports is connected via a FET with a 5 to 6 ohm Ron resistance. The second chip mentions "5V tolerant I/Os", the first doesn't. According to TI, and I know they're talking about CB3T not CB3Q, but that section is about FET switch translators in general:

Devices from TI’s CBT, CBTD, CB3T and TVC families can be used in level-translation applications. The diagram shows a CB3T bus switch interfacing a 3-V bus with a 5-V (TTL) bus. The CB3T device down-translates the signals from the 5-V bus to 3-V levels. No translation is necessary to transfer signals from the 3-V bus to the 5-V (TTL) bus, since the VOH level from the CB3T switch is greater than the required VIL of the 5-V (TTL) devices connected to the 5-V bus

Also note that if you install the CB3Q on that wing, you will be powering it with about 4.3V after the diode drop whereas its nominal rating is 2.3v - 3.6v with absolute maximum rating of 4.6v.

What you're probably really after is something like dual supply level translators such as SN74ALVC164245 or SN74LVCH16T245. Note in their datasheets that one side of the I/Os is powered by VCCa while the other side of the I/Os is powered by VCCb) where one of these supplies will be 3.3V and the other will be 5V.

[neslekkim]

woha, I havent gotten around to use it yet, was planning to setup the logic analyzer software on the papilio one, and use it.

Does this mean that all these bufferboards is not good, or just some batches?

[madscifi]

Anything with a chip labelled CB3Q16245 will not work correctly. Based on the spec sheet I believe that boards with the IDT chip will also fail, but I don't have a board with that chip to test and so it is possible I'm wrong about that.

This issue only affects the 16-Bit I/O Buffer Wing boards. The Logic Sniffer 16-bit Input Buffer Wing is not affected.

If someone can test one of the boards with the IDT chips, please make certain the board is not attached to a Papillo or OLS (or any other device) when the test is made - the internal diodes on the pins of the Xilinx chip will confuse the issue.

[neslekkim]

The one I have, an Buffer wing, BPW5015, bought from this shop, have the CB3Q16245 chip.. 

Should I discard it then?

 

Btw, Should I not use that type of wing with the OLS?, is it supposed to be only the Input type?, I bought this one because I thought it would be more versatile..

[Jack Gassett]

Neslekkim,

 

I have the new design on my desk now, the only problem is that I have a deadline to complete a presentation for Design West by 3/22. It is eating up all of my time right now, but once I finish it those boards and the next LogicStart board are the top priority. I will hopefully be able to start swapping out these boards towards the end of March.

 

Sorry for the long wait.

Jack.

[neslekkim]

ah, good, I see now that I didnt read eveything fully, that you did redesign it, I thought it just needed another chip.

So I send an email then, to get on the list for new version?

 

On your site, you differ between Input buffer and Buffer wing, I bought the one called: "16-bit I/O Buffer Wing Item# BPW5015-IO", but I see that Seeedstudio have this on stock:

http://www.seeedstudio.com/depot/logic-sniffer-16bit-input-buffer-wing-p-721.html?cPath=174

With another chip, and calls it Input buffer wing, but it is the same pcb (albeit red), I guess that one will do fine as OLS wing, until one gets the Buffer version correct?

[MartyMacGyver]

I was *just* now pondering ordering one of these boards myself... I take it that it would be a good idea to hold off on buying this elsewhere (I thought it was strange I couldn't find it here; now I understand).

[neslekkim]

if you only need input, you can buy this: 

http://www.seeedstudio.com/depot/logic-sniffer-16bit-input-buffer-wing-p-721.html?cPath=174

 

and these probes:

http://www.seeedstudio.com/depot/open-logic-sniffer-probe-cable-p-619.html?cPath=178_180

the probes are the cheap ones..

I guess these would work as well:

http://www.saleae.com/accessories

[MartyMacGyver]

I found the probes on CuteDigi as well, where I'm ordering a few other small components anyway.

 

Speaking about the Buffer Wings though, I'd kinda rather get the I/O (seems the specs on the I/O board are better - assuming it wouldn't cause problems of its own during logic probing by virtue of it having output capability).

 

Actually, it's a little unclear how this I/O board is supposed to work (even looking at the schematics I found elsewhere on here). Does it auto-sense data directions or does it steam some I/O lines to do that?

 

To recap though, the current input board available at seeedstudio is fine (is it available directly from here and I missed it?), and it's the I/O board that I may not really need which has some work being done on it but should be ready soon. (Edit: and I won't get the I/O wing at CuteDigi either since it sounds like it'd have the same problem - too bad they don't just have the input board.)

[Jack Gassett]

I found the probes on CuteDigi as well, where I'm ordering a few other small components anyway.

 

Speaking about the Buffer Wings though, I'd kinda rather get the I/O (seems the specs on the I/O board are better - assuming it wouldn't cause problems of its own during logic probing by virtue of it having output capability).

 

Actually, it's a little unclear how this I/O board is supposed to work (even looking at the schematics I found elsewhere on here). Does it auto-sense data directions or does it steam some I/O lines to do that?

The I/O board uses a bus switch chip that does not do voltage translation like most chips do. Instead it acts as a FET switch that is an actual physical bi-directional connection between the pins. By setting the FET voltage to 4.3V we get a voltage drop of 1V which means that the FET will never allow more then 3.3V to be output on any of the pins. If you input 5 volts on either side it will restrict the output on the other side to 3.3V. If you input 2.5V on one side it will output 2.5V on the other side.

 

The benefit of this is that it is bi-directional but also very, very fast. It can operate at speeds up to 2GHz.

 

I should be able to get the chip tested this week and make sure it works as intended. I made the mistake of allowing a different chip to be used on the old design, the original became unavailable, and did not realize it was not working as intended. This new design and chip should work as intended and I should be able to get this board available again soon.

 

 

To recap though, the current input board available at seeedstudio is fine (is it available directly from here and I missed it?), and it's the I/O board that I may not really need which has some work being done on it but should be ready soon. (Edit: and I won't get the I/O wing at CuteDigi either since it sounds like it'd have the same problem - too bad they don't just have the input board.)

[neslekkim]

No news about the io buffer replacement yet?

[Jack Gassett]

I tested the new design and it works as expected, I wanted to manufacture a batch myself but that just hasn't been possible. I'm dealing with a family medical emergency right now and will be out all next week for Design West. After that I will either initiate a batch to be manufactured or do some myself, I have all of the components to do so but making the time is the difficult part.

 

Jack.

[neslekkim]

Sorry to hear that, hope everything goes fine.

 

I was just checking in, no hurry..

[neslekkim]

Any news?

[Jack Gassett]

I just built 25 of the new boards, will have to test them next and ship them to Kalesh in Oregon. We should be able to replace your board in the next couple days.

 

Jack.

[neslekkim]

Good, give me an option to buy an extra at the same time, and to pay some for the replacement, I think sending it from Norway to US can be expensive/interesting..

[neslekkim]

Just discovered that the wing was in the store now, so assuming it is the correct one, I ordered a couple to replace the defunct one.

according to picture and productnumber it seems like the same I already have, does that mean I could have gotten an chip, and only replaced that on my wing here?

 

I thought there was some other changes beeing done also?

You mentioned 2ghz above, but on the productpage it is 200mhz? not thinking I need it but what is correct?

[Jack Gassett]

Kalesh was just getting the page up, it was just a copy from the old page and was waiting for me to verify and update it.

 

 

Just discovered that the wing was in the store now, so assuming it is the correct one, I ordered a couple to replace the defunct one.

according to picture and productnumber it seems like the same I already have, does that mean I could have gotten an chip, and only replaced that on my wing here?

 

I thought there was some other changes beeing done also?

You mentioned 2ghz above, but on the productpage it is 200mhz? not thinking I need it but what is correct?

Its definitely a completely different design for a 5V part instead of a 3.3V part. There is a new diode that you should be able to see that sets the limiting voltage to 3.3V. So it would not be possible to just replace the chip for this one.

 

200Mhz was the number from the Input only buffer, it was incorrect for this new version and is removed now. However, the datasheet for the chip we are using now does not specify a speed, many similar chips say 2Ghz, so I'm sure its above 200Mhz, but without testing or a number from the datasheet I don't want to make any claims.

 

Jack.

[neslekkim]

ok, it will definately be fast enough for my use anyway, just wanted to ask since it was written something on the page, and you wrote something else before in this thread.

Ordered a couple of them, hope I can start playing with my p* soon

[MartyMacGyver]

Thanks for ironing out the bug in the original version... just ordered one myself. :-)

[Jack Gassett]

If anyone wants a replacement for your Buffer I/O Wing board please send us an email to support@gadgetfactory.net.

 

Thanks,

Jack.