13

I recently got a SID chip (8580r5) for $35 and got it playing music. I was really happy about it. I hooked up a small amp chip to the 9V supply momentarily (it didn't work, I was building that part of the circuit) and after a second of disconnecting it the SID stopped playing music and I quickly saw it was drawing a lot of amps on the 9V rail, I think around 600mA when it usually draws around 10mA. I quickly shut everything down (I think within 1 to 2 seconds of it going bad) and the chip had gotten a little bit warm. Now if I bring up the 9V supply slowly the current increases relative to the voltage which is not what should happen, it should stay at around 10mA. It draws the same amount of current on the 5V supply as it did before but if I quickly supply 9V too it doesn't produce any output.

How did I break it as the only connection the amp circuit had with the SID was through a 2.2uF cap on its output and a common ground?

I took extreme care to not have a SID chip die on my hands

I'm hoping that if I let it sit for a moment it will be magically fixed but I don't like my chances.

Here is a small schematic of what I had. I couldn't find a TDA1524 and my circuit is the same as the one on the datasheet except the components for the right channel where missing and some values where a little bit different so I didn't bother drawing it, tell me if you want me to.

enter image description here

EDIT: if i don't connect 9V i can pick up a 20mV signal of the normal output of the SID so only an output pre amp died most likely explained by @Justme's answer https://www.bitchute.com/video/xo4R3Na8VDCH

| improve this question | | | | |
  • What "small amp chip" did you use, and how did you connect it? Can you draw a circuit diagram? – Michael Graf Mar 14 at 17:02
  • @MichaelGraf I used the TDA1524, the circuit i had was the same on the datasheet except all the components for the right channel where missing and some values where not exactly what the datasheet called for. I have since completely removed it to test the SID but even if i had completely messed up the circuit i can't imagine how it could hurt the SID – user16542 Mar 14 at 17:10
  • The datasheet I'm looking at (the Philips/NXP one at pdf1.alldatasheet.com/datasheet-pdf/view/19211/PHILIPS/…) doesn't even have an application circuit. If you want others to do more than just guess randomly what might have happened, please provide a circuit diagram showing the power supply, the SID, the TDA1524 and all the components directly connected to them. Sorry to insist, but there's very little to go on without that. – Michael Graf Mar 14 at 17:29
  • @MichaelGraf Sorry, I understand. If you want I can draw a schematic on kicad but the application circuit is on the 3rd page of the PDF you linked, It looks like a block diagram but it's both. the sid was controlled by an Arduino and the power supply was the Arduino for the 5V and linear bench power supply for the 9V. – user16542 Mar 14 at 17:36
  • Sorry, you're right about the datasheet. Looking through the datasheet, I thought that the schematic on page 3 contained only the internal circuit, and overlooked the external components. – Michael Graf Mar 15 at 11:00
16

The amplifier input is biased to about half supply voltage. The SID output is also biased to about half supply voltage. You have little or no bulk/bypass capacitance on the 9V supply pins, so abruptly disconnecting the 9V supply from amp also abruptly shuts down the amp input bias. But as you have the quite excessively large 22uF coupling capacitor between SID output and amp input, the quickly falling amp bias also pulled down SID output quickly down, and depending on actual bias voltages, perhaps it took SID output negative which damaged it. Sensitive chips should be protected by buffering with local power supply - I mean if you made a SID device, you would not connect the SID output directly to outside world either, so same in this case - do not rewire things while powered.

| improve this answer | | | | |
  • Thanks a lot, that makes more sense. So this could have been prevented if I had a local supply on the amp chip in the form of a capacitor on its power pins so the amp's bias falls slowly and doesn't bring the SIDs output too low right? – user16542 Mar 14 at 20:07
  • 2
    I meant you should have a buffer op-amp near the SID which uses same 9V than the SID does. When the op-amp is the one that is always powered along the SID and being connected to outside world like amp on another board with glitchy power supplied, the op-amp can take more abuse. Abruptly giving 9V to amp can also make the input bias to abruptly rise to 4.5V, so I really don't recommend toggling amp power on or off while being connected to SID. – Justme Mar 14 at 20:16
  • 2
    I certainly wouldn't suggest to reproduce it on an actual poor SID :), but you might test the assumption mocking the SID with some passives, power TDA off abruptly and look at the transient capture. Can just add that on the C64, the output is just buffered by a suitably biased NPN transistor and then AC coupled to the outside world through a 10 uF cap. – Retrograde Mar 14 at 20:59
  • 1
    IT'S ALIVE (sort of). @Justme YOU ARE RIGHT!. i plugged it back in the circuit and of-course it didn't work and drew 900ma on 9V but if i don't connect the 9V i can see a 20mV signal on the output on my oscilloscope and if i amplify it it's the sound from the SID. The chip works fine but what i assume is an output pre amp is dead and the only way that can happen is with what Justme said. I can still kinda hear my SID playing tunes :) ! – user16542 Mar 14 at 21:13
  • 1
    @Retrograde i just noticed it only needs an atmega8515 which costs about $1 so i'll give it a try :) Also i recorded a video of the dead SID playing commando.sid bitchute.com/video/xo4R3Na8VDCH :) – user16542 Mar 14 at 23:31
2

This sounds like an ESD/over voltage event might have occurred while you were working with the circuit. The SIDs are extremely sensitive.

It is not uncommon for a defective SID to draw excessive current.

(This is not going to bring back your SID, but just a tip for future work: while prototyping your project, consider using a SwinSID (Nano) instead - it's emulation but pin compatible, cheap, harder to damage and inducing less sorrow when damaged.)

| improve this answer | | | | |
  • I hadn't touch the chip, It was well hidden below a bunch of wires which i also didn't touch, I briefly touched the amp chip's power pin to the crocodile clip of the 9V supply and a second after releasing it the sound quickly faded to nothing and the chip started drawing a lot more current – user16542 Mar 14 at 17:13
  • 2
    If the SID shared power rails with the TDA, that would still make the SID susceptible to an ESD event when touching the TDA part of the circuit. Is the supply regulated? – Retrograde Mar 14 at 17:20
  • yeah, I got 5V from the Arduino and 9V from a linear bench power supply. The moment it died i think i was holding on one hand the 9V crocodile clip from the plastic and the power for the TDA on the other, I hadn't moved for hours. I don't think I was statically charged. Could it be something else? – user16542 Mar 14 at 17:27
  • 1
    If there is any significant inductance in the power distribution a sudden current surge can cause high voltages to appear on the power line that can damage any attached devices. It is what is called "load dump" in cars. Also you should NEVER change your wiring (especially power) while your circuit is running unless you are very experienced. – Kevin White Mar 14 at 17:48
  • 1
    While the SID is quite sensitive, and ESD can happen and it is often used as an explanation for all kinds of sudden mysterious failures. But based on the description on the events and by looking at the schematics and datasheets, I believe it is far more likely that it got destroyed when unpowering the amp chip abruptly while being connected to the SID chip. – Justme Mar 14 at 19:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.