ARP 1601 clone Build Thread

[elmegil]

Have you double checked that Q4/Q5 are the right types of transistors? Q4 is 3906, Q5 is 3904.... Having those reversed, or one of them wrong could stop the draining from happening.

[fallout]

Yes, it looks like C7 drains to ground through Q5 but only happens manually (pressing button). I'd assume this rules out Q4/Q5?

What's responsible for creating the continuous cycle of drain/charge is what I don't understand.

[fallout]

Have you double checked that Q4/Q5 are the right types of transistors? Q4 is 3906, Q5 is 3904.... Having those reversed, or one of them wrong could stop the draining from happening.

Yes, that was one of the first things I checked.. Like I mentioned above, I figured these were working since C7 does in fact drain when start/stop is pushed.

[elmegil]

I apologize I wasn't clear that you'd checked or that you see it drain.

Can you explain a little further though? When you say you see it drain, you're seeing the saw down slope one time? I would guess only when it's in Gate mode?

[Kipling]

There have been some issues recently with certain LM301AN chips, so you may want to try an alternative if you can.

[fallout]

I apologize I wasn't clear that you'd checked or that you see it drain.

Can you explain a little further though? When you say you see it drain, you're seeing the saw down slope one time? I would guess only when it's in Gate mode?

Scope on TP4 and in gate mode, it appears that the capacitor drains from +15v to 0 when holding the start/stop button. The rate at which it drains depends on the position of the Clock Frequency slider (lowest position = slow drain, highest = quick). Makes sense.

So here's my thought after reading the service manual a few times. It looks like the comparator is what recharges C7 back to 15v after it drains to 7.5v. It looks to me like that part of the circuit is where the problem is. For instance, I'm assuming when in 'trigger' mode (or in gate mode when holding the stop/start button) and the sequencer is running, C7 is constantly being drained and once it reaches the 7.5v threshold, the output of the comparator (by way of the one shot circuit) charges it back to 15v. This repeats until you press stop or in the case of gate mode, release the button (turning Q6 on/off?). It's obvious that C7 isn't being charged back up when it reaches the 7.5v threshold and just drains to 0v. Is my thinking correct here?

Obviously, I will first replace the 10n currently in C7 with the correct .1uf cap but as you mentioned earlier, it doesnt seem like that would be the issue. I have a funny feeling about the CA3086 in Z5 and will replace that as well and see where that gets me.

While I wait for parts, Is there a way to diagnose the comparator circuit? I've already checked the resistor values in the comparator circuit and they look fine. I also used my multimeter's diode test to try and test the transistors on the CA3086/Z5 and am getting some wonky results. Not sure if it's because I'm testing it in circuit but results seem a bit funny. For instance, pin 2 -> 1 and 2 -> 3 are both reading open (tried reversing test leads as well)

Removing an SOIC should be fun! Ha!

There have been some issues recently with certain LM301AN chips, so you may want to try an alternative if you can.

I tried this and it didn't seem to help. Thank you for the suggestion!

[elmegil]

So yeah, if you can observe the drain while you hold the switch down in gate mode, AND it goes all the way to zero, then something is not resetting Q6. That might be that the comparator isn't firing (seems most likely at this moment) at 7.5V, it might be that some other part between there and going back to turning on Q6 to recharge isn't working. I do recall catching that reset pulse from the comparator was pretty hard, but if you have a scope and set it for "single sweep" triggered on that output going high, you ought to be able to catch it, if it happens. The 3086 becomes a high suspect. If you don't have a spare, a 3046 will do in a pinch.

[elmegil]

Oh, if you have the SOIC 3086 ... you're probably going to want hot air to get that bad boy off there.

There are some other alternatives though... You can heat one side with solder braid. They make tools specifically for this which are like heated ... not pliers but long blades that heat up and let you adjust where they go, so you can heat both sides at the same time. You can try dumping a bunch of solder onto one side at a time, so the solder carries the heat across all the pins.

If you don't want to save the chip, just carefully cut all the legs off and then take the legs out one by one with solder braid.

[fallout]

Oh, if you have the SOIC 3086 ... you're probably going to want hot air to get that bad boy off there.

There are some other alternatives though... You can heat one side with solder braid. They make tools specifically for this which are like heated ... not pliers but long blades that heat up and let you adjust where they go, so you can heat both sides at the same time. You can try dumping a bunch of solder onto one side at a time, so the solder carries the heat across all the pins.

If you don't want to save the chip, just carefully cut all the legs off and then take the legs out one by one with solder braid.

I already got the chip off the PCB fairly easily. I wasn't concerned with saving the part, so I flooded the chip with flux and applied a blob of solder to each side and gently grabbed the chip with tweezers. Similar to your suggestion. Cooks the chip but works like a charm!

[fallout]

Does anyone in the US have a few, or know where to get a few CA3086 SOIC-14's? Synthcube sent replacement parts but inadvertently sent DIP14's instead. I'm sure they'll send the right part but I was looking for someone/somewhere else to get a few in the interest of time. The CA3086 SOIC-14's seems to be more rare than the DIP-14. Thanks!

[LED-man]

Please look for: LM3046M

[Kipling]

There is also the AS3046 now, in DIP and SOIC.

[fallout]

Thanks guys.. The LM3046M isn't stocked many places either but does seem to be more available than the 3086. Is there a tradeoff for using a 3046 instead?

[Kipling]

It was always a 3046 chip anyway. A 3086 would probably work but they were discontinued years before the 3046 was, so will be harder to find. I’m not actually sure the 3086 was ever produced in SOIC.