Hi.
I have an Arcam Delta 290 that I've just recapped and I've found some other problem too. Some resistor broken... I broke the copper rivets that connect the filter capacitors to the back side of the PCB, as I'm a beginner with diy soldering and desoldering, and I'm waiting for the spare rivets to arrive. Actually there's something more but I will tell in another moment.
In the meantime, I wanted to learn how to use LTSPice and I thought starting with the mains and power stage I'm working on would be good as I could get the several voltages for checking.
This post of @Mooly got me started but I'm drawing it trying to keep it as much similar as possible to the service manual, even because I'm not that good in electronics and I can't replace branches of the diagram with equivalent circuits.
Other than still needing to find LTSpice models for certain components, I'm wondering how the crossing wires of the diagrams should be considered: connected/not connected? Should you Mooly read this, I'd really be grateful if you told me how you did discern the touching from the non touching crossing wires.
I gave the variable resistor RV1 a fixed value that I change time by time to test the IQ across R1 but that voltage doesn't change, therefore something must be wrong.
Thank you!
![CropperCapture[7].jpg](https://www.diyaudio.com/community/attachments/croppercapture-7-jpg.1238398/ "CropperCapture[7].jpg")
I have an Arcam Delta 290 that I've just recapped and I've found some other problem too. Some resistor broken... I broke the copper rivets that connect the filter capacitors to the back side of the PCB, as I'm a beginner with diy soldering and desoldering, and I'm waiting for the spare rivets to arrive. Actually there's something more but I will tell in another moment.
In the meantime, I wanted to learn how to use LTSPice and I thought starting with the mains and power stage I'm working on would be good as I could get the several voltages for checking.
This post of @Mooly got me started but I'm drawing it trying to keep it as much similar as possible to the service manual, even because I'm not that good in electronics and I can't replace branches of the diagram with equivalent circuits.
Other than still needing to find LTSpice models for certain components, I'm wondering how the crossing wires of the diagrams should be considered: connected/not connected? Should you Mooly read this, I'd really be grateful if you told me how you did discern the touching from the non touching crossing wires.
I gave the variable resistor RV1 a fixed value that I change time by time to test the IQ across R1 but that voltage doesn't change, therefore something must be wrong.
Thank you!
I haven't attempted to compare your LTSpice schematic with the service manual (you might post it for convenience) but I see a couple of errors. R30 and R31 can't be correct--- likely 100K as a guess. I think there's a missing wire between Q14 emitter and Q7 collector. There may be other errors I've not spotted.
You'll be in good hand with Mooly. 🙂
You'll be in good hand with Mooly. 🙂
Thanks BSST! You're right! I've in the meantime found I used "," instead of "." in some resistance values too. Still tiny V drop across R1 with varying RV1
Do you think the base of Q14 should be connected between RV1 and R17? I tried with not much difference.
Here is the LTS schematic
Do you think the base of Q14 should be connected between RV1 and R17? I tried with not much difference.
Here is the LTS schematic
Last edited:
Sorry, I skipped the missing wire. I'm doing the change and uploading again in a while
Edit: big V drop now but the wiring must be quite messed up, I guess. Updated the LTS file
Edit: big V drop now but the wiring must be quite messed up, I guess. Updated the LTS file
You are missing a supply for the opamp and also try a an LT1056 which is a FET opamp like a TL071.
I can't believe I didn't keep the sim I did earlier. Can't find it on this PC...
I can't believe I didn't keep the sim I did earlier. Can't find it on this PC...
Thank you Mooly!
I did as you said. Still 4 and something volts across R1 despite the RV1 value, unfortunately.
Do you mean I should install the 2023 LTSpice?
I did as you said. Still 4 and something volts across R1 despite the RV1 value, unfortunately.
Do you mean I should install the 2023 LTSpice?
I see what you mean about the voltage. I'd have to go through the outputs stage and compare it with the Arcam circuit.
That is just a link in my signature line 🙂
I've seen now 🙂
Yes, 0 bias even without snipping C3, here. What about the wiring? I can't understand when I should consider two crossing wires as a node and when not.
Thank you for your effort, Mooly
You have to rely on your own theory of circuit design 🙂
More... the little FET CCS looked wrong, way to much current, 1ma is more typical for an input stage. R16 is shown as 82k on the diagram.
This sets OK now if you play with the bias. Alter R17 to do that for now. 5600 ohm brings it down to around 4ma
Hi Mooly.
I wish I was that good. I'm good in IT, much less in electronics. I only studied electrics (I don't know if the english is right), not electronics and it was long time ago!
Actually yes, I checked on the amplifier and the actual R16 is not 2K but it's 56K.
I will try as you say. Though the actual circuit lets me adjust (only a channel at the moment) the bias turming by a small fraction of range the RV1 trim pot, which is a 1KOhm one.
I will dig into this. Thank you very much
I wish I was that good. I'm good in IT, much less in electronics. I only studied electrics (I don't know if the english is right), not electronics and it was long time ago!
Actually yes, I checked on the amplifier and the actual R16 is not 2K but it's 56K.
I will try as you say. Though the actual circuit lets me adjust (only a channel at the moment) the bias turming by a small fraction of range the RV1 trim pot, which is a 1KOhm one.
I will dig into this. Thank you very much
The bias range on the real thing will be pretty limited and very dependent on the actual characteristics of the devices used. If you have replaced transistors in the output stage and particularly the FET's then I would think you would have to tweak the bias setting resistors to get the preset back in range.
The actual problem now: on one channel, turning the trim pot to adjust the bias, nothing changes. Alwas 0 V across R1!
I took the pot apart and checked it and it's good. I took several transistors apart and they're working 🙁
I took the pot apart and checked it and it's good. I took several transistors apart and they're working 🙁
Sorry i posted without seeing your response. Yeah, I understand what you mean. Anyway I haven't replaced anything except a resistor.
For my other post, I'm editing it as I forgot to mention the damage I did
As I can't edit it, I needed to say i damaged the contacts in the PCB and then i will try to simulate the damage. Can it be the filter capacitors not doing contact to both sides of the pcb prevent me adjusting the bias and getting the above said 0 V across R1? Not home now and I will try later
You need to measure the voltage between these two points as you turn the preset and the voltage should be altering as you turn the preset. As the voltage increase the bias current should also increase.
You must measure between the points. This is biased at 33ma (3.6mv across your resistor) and shows typical voltages So you would measure 7.41 between these points. These will be different in a real amp of course and different for differences in semiconductor characteristics but they are in the right ball park.
And at 'just' zero bias current in the output FET's. So 5.44 volts now.
And at a forced 100% zero bias (R17 cut out). Now 1.63 volts measured.
Only the big reservoir caps being disconnected would cause big problems. If you mean smaller decoupling caps then they should not alter the DC conditions.
Hi Mooly.
Last night I was digging into what you suggested. That's really helpful, thank you very much.
I've found R17 to actually be 3.1 KOhms. Replacing it in LTS, it gives really little range of variation around R1. What do you think?
The resistors look to be the original ones, as I can recognise resoldered ones.
At this point, do you think I shoud replace R16 to 82K (as per your test) and R17 to 4700 Ohms as per the original project?
I will check all resistors now to see the "actual" values.
I meant the big caps. I only checked the amp with the bulb test. Anyway, that not being the culprit makes sense because those caps are not channel related.
Now I need to buy some resistors and I'm still waiting for the copper eyelets to restore the front to back pcb contacts. Then I'll test on the real amp
Last night I was digging into what you suggested. That's really helpful, thank you very much.
I've found R17 to actually be 3.1 KOhms. Replacing it in LTS, it gives really little range of variation around R1. What do you think?
The resistors look to be the original ones, as I can recognise resoldered ones.
At this point, do you think I shoud replace R16 to 82K (as per your test) and R17 to 4700 Ohms as per the original project?
I will check all resistors now to see the "actual" values.
I meant the big caps. I only checked the amp with the bulb test. Anyway, that not being the culprit makes sense because those caps are not channel related.
Now I need to buy some resistors and I'm still waiting for the copper eyelets to restore the front to back pcb contacts. Then I'll test on the real amp
Attachments
Last edited:
I think the original design values will be tightly tied in to very specific semiconductors bought in by Arcam.
The actual resistor values are not very critical as long as you get the adjustment in range. Making the preset higher in value would give you more range. You could try a 1k or 2k and/or tweak the resistor values. We can't calculate this because the actual semiconductor turn on voltages are not known accurately. The FET's are probably where the big variations are tbh.
Lowering R17 will increase the bias current and increasing R16 will do the same. Its trial and error in the real thing, particularly with such a low value preset. R55 is present to force zero bias should the preset go open circuit. It does that by fully turning on Q14
Hi Mooly.
Still waiting for some spares here. The bad of living in remote areas...
In the meantime, I did some check and pulled TCR504's (J504 jfet's) from both channels.
I used a 100 Ohm resistor in series and tested the current with two different V sources:
Don't you think they are faulty?
Edit: I was forgetting: R23 is 18K and not 10K. On the diagram, it is not much readable
Still waiting for some spares here. The bad of living in remote areas...
In the meantime, I did some check and pulled TCR504's (J504 jfet's) from both channels.
I used a 100 Ohm resistor in series and tested the current with two different V sources:
- 12V DC
- channel 1: 93 mA direct; 0.9 mA reverse
- channel 2: 93 mA direct; 0.9 mA reverse
- 1.2V DC
- channel 1: 3.7 mA direct; 0.7 mA reverse
- channel 2: 2.4 mA direct; 0.7 reverse
Don't you think they are faulty?
Edit: I was forgetting: R23 is 18K and not 10K. On the diagram, it is not much readable
Last edited:
It looks like you are getting the right results to me 🙂
0.9ma looks correct and the 93ma would be what you would see if you applied 12 volts via a 100 ohm with the FET reversed. So your readings are correct but the interpretation wrong if that makes sense. The absolute max gate current is 20ma so you have exceeded that many times over...
The gate to source and gate to drain behaves a bit like a diode when forward biased.
0.9ma looks correct and the 93ma would be what you would see if you applied 12 volts via a 100 ohm with the FET reversed. So your readings are correct but the interpretation wrong if that makes sense. The absolute max gate current is 20ma so you have exceeded that many times over...
The gate to source and gate to drain behaves a bit like a diode when forward biased.