Help with SS Tarantula 800/5

[jacampb2]

Hi, I am hoping someone can give me a hand with this thing. I bought this 800/5 in non working condition simply because I have a bit of a weakness for the SS tarantulas. I buy and repair these and PG amps just as a hobby. I have a sizable collection going, and pretty much every one of the amps have been purchased in similar condition. This is the first one that has beat me up quite this badly and I could really use some help.

Anyhow, here is the deal. When I got the amp, it had no PS fets in place for either supply. It is also missing all the PS fet spring clamps, but that is another issue-- I found the PNP halves of the gate drivers were blown, some of the gate resistors open, and in the end, the SG3525 itself was bad. After rebuilding all of this, the PS would come up for less than 3 seconds, then the PS section would emit an audible squeal and it shut back down. I have had amps squeal like that when the PS was under heavy load, so initially I pulled all the outputs to check for shorted or leaky BJTs (I had checked the transistors while in circuit, and all looked okay, but I wanted to be thorough). In the end there were no issues, so I moved on.

Eventually, I found that I still had gate drive pulse while the supply was apparently shut down, but the duty cycle was so low that it was virtually non existent. Further trouble shooting showed that the inverting input of the error amp on the SG3525 was much higher than the non inverting input. The compensation pin was at nearly 6Vdc throughout it all. Anyhow, I finally temporarily bypassed the inverting input, and gave it a lower voltage than the non inverting input and the amp powered up and worked perfect.

With the error amp on the SG3525 effectively bypassed, the amp behaves perfectly. From what I can tell the rails are fine, the 4 channel section has +/-35Vdc, the mono channel has +/-50Vdc, the amplifier sounds fine on all channels, it does not appear to have any problems left except this one with the PS.

I have been trying to determine what exactly they were doing with this, it appears that the comp pin is driven directly, but the error amp is also in use. I was under the impression that it needed to be either/or, but it seems I am mistaken. I was hopeful that SS would give me the schematics, they have been very helpful with prints for a few other of their other amps I have worked on, but once I reached them, they told me they are no longer allowed to distribute this information.

I'm hopeful that someone on here may have worked on one of these, or may recognize the design and be able to provide some assistance. Thank you in advance! Bellow is a list of the pin voltages on the SG3525, and then some pictures of the amplifier.

SG3525 pinouts:
1) 2.554 v
2) 4.84 v
3) 13 mv
4) 201.6 mv
5) 2.094 v
6) 3.80 v
7) 6.2 mv
8) 4.87 v
9) 5.76 v
10) .517 v
11) 82.4 mv
12) 4.4 mv
13) 11.36 v
14) 85.5 mv
15) 13.57 v
16) 5.14 v

Here are some pics:


Click for Full Size



Click for Full Size



Click for Full Size



Click for Full Size

Thank you all in advance. Any help will be appreciated.

Sincerely,
Jason

[Perry Babin]

Are you sure that pin 7 is correct? It's generally very nearly the same as pin 5.

I have information on a similar amp. Email me if you want it.

babin_perry@yahoo.com

[jacampb2]

I'll check, maybe I didn't get a good connection with my probe. It will probably be tomorrow sometime before I get some free time, but thank you for the offer to help. I'll drop you an email as well with a request for that info Thanks!

Sincerely,
Jason

[jacampb2]

I got a quick recheck of the PWM controller voltage's and the above was incorrect. I do have 2.066 v on pin 7, I don't know where I got the 6.2 mv before, I must have been probing air

That's probably all I am going to get to work on it for the next few days, I am heading back to work tonight-- 12 hours midnight shifts for the next four nights.

I'm thinking I have been going about this the wrong direction though. I have been thinking that the inverting input is to high, but on further study of the information Perry provided for me, I think that it may actually be the non-inverting input is being held to low! That will be a whole different plan of attack, but I will get back to it next week and keep you guys updated as to the end result.

Thanks,
Jason

[jacampb2]

I have been studying my little problem in my spare time here at work tonight, and trying to work out the solution, and I just realized that I have been confused about which input is which on the SG3525's error amp. I was thinking that Pin 1 was the non inverting input and pin 2 was the inverting input. I was incorrect, it is exactly opposite. Pin 1 is inverting, two is non inverting. The pin I actually pulled low when I got this thing to work was pin 2. I also did not pull it all the way to ground, but rather split the 5v reference and applied this to pin 2, that puts both inputs very close to the same, but IIRC, pin 2 was still the lower one after doing this. Anyhow, long story short, the references I made above were backwards, which brings us to a new question.

Shouldn't the supply be working with the voltages listed above? If I understand correctly, the inverting input has to be lower than the non inverting one, and comp has to be high, above 5V for full duty cycle, and the IC should run wide open...

I've almost confused myself more than when I started asking questions But that is the only way to learn, right?

Thanks,
Jason

[Perry Babin]

With everything as it normally would be (remove the modifications you made), what happens to the output (pins 11 and 14) if you connect pin 10 to pin 12?

[jacampb2]

Yeah, that was the only one that was standing out to me too. That is the shutdown pin, correct? Are you thinking that even with it bellow .6v it is still having an impact?

I already removed the modifications, the voltages I listed above are all bone stock as it came to me-- I'll definitely pull pin 10 to ground and see what happens. Hopefully I will get up early enough tomorrow to tinker with it a bit, but it's probably unlikely.

Thanks for the help!
Jason

[Perry Babin]

I've seen a few ICs that were leaking internally and wouldn't produce output when that pin was not nearer to ground. For normally functioning ICs, I think 0.6v is the threshold (for the ICs that I tested).

[jacampb2]

Well, I got my butt out of bed early enough today and tinkered with the amp for a while. Pulling pin 10 to ground does indeed make the supply come up. Pins 11 and 14 are showing ~4.62Vdc each and the rails and auxiliary supply all come up and test normally.

So, where to go from here. Pin 10 is driven by a TL072 op amp, The output of the TL072 goes through a series diode and a series 100k resistor. The output of the op amp is at 1.36V, after the diode it is ~1.3v, after the resistor it is the same as at pin 10, ~.51V. The inverting input of the op amp is tied directly to VRef on the SG3525, so it is at 5.14V, the non inverting input is tied to several other components, a small signal transistor and apparently a 10K pull down resistor tied directly to ground. I believe there are other components also, but I have not yet had a chance to track them down. The non inverting input has approximately 1.6Vdc on it.

I swapped out the op amp because I broke a pin on it when I was trying to remove it to track down some traces routed bellow it. The new op amp has no change from the previous one.

On a side note, does anyone know what components labled TH on the silk screen might be? I had guess that it was for "thermistor" however, while I was trying to track down all the locations that Vref goes, I discovered one of these components marked TH with Vref on one end of it and a few hundred millivolts on the other. This particular component looks like a glass body diode, but it does not have the dark band, it more looks like a glass tube with two tight coils of copper in it. The other components marked TH on the board look more like a traditional thermistor, but none of these seem to be close enough to the sink and discrete components to provide reliable thermal protection. I know that there are methods for sensing temperature with a diode, but I am not really familiar with it, and I guess I would have thought it would be marked differently. You can see the part in question right next to the fans power supply header in my 4th picture, open the full size link and it is pretty clear.

Also, in that same picture, both above and bellow the rail caps there is a long line of devices that look a lot like normal switching diodes, but again are not banded and the case is a bit lighter gray then a normal diode. These guys are all labeled Bxx on the silk screen. I have never run across these before either, I don't suspect an issue with them, I am just curious as to what they are.

Thanks again for all the help!

Later,
Jason

[Perry Babin]

The orange device is almost certainly a thermistor.

The gray components are likely ferrite beads.

Replace the 3525.