problem is in assymetry you have with feed of opt bjt
it's fed from positive rail and when you have offset, there is lesser voltage across 1K5, so Iq is different; if there is CCS instead of 1K5, Iq will be stable, but that will be just band-aid, not initial problem solved
initial problem - try to re-do your servo, to keep offset from cold, and then Iq will be stable too
just glancing at your circ - try lowering value of 150K
you need to meet ne condition - Iq of Q19 CCS need to be lower than it's needed for proper operation of Q17/Q21 LTP, so servo Ic is called to compensate "missing" current
fact that you initial DC Offset is positive says that IRF610 are not open enough, which means LTP is not conducting enough current
that's where servo is called
it's fed from positive rail and when you have offset, there is lesser voltage across 1K5, so Iq is different; if there is CCS instead of 1K5, Iq will be stable, but that will be just band-aid, not initial problem solved
initial problem - try to re-do your servo, to keep offset from cold, and then Iq will be stable too
just glancing at your circ - try lowering value of 150K
you need to meet ne condition - Iq of Q19 CCS need to be lower than it's needed for proper operation of Q17/Q21 LTP, so servo Ic is called to compensate "missing" current
fact that you initial DC Offset is positive says that IRF610 are not open enough, which means LTP is not conducting enough current
that's where servo is called
Yes, you are right, didn't think about that.
Ahaaa, thanks, I will play with that. I didn't know that.
Thanks ZM, lots of inspiration to be busy with coming days
It is OK now, first put in a 2K2 resistor instead of the 150K --> 'oscillating' like hell....you should have heard the noise coming from my old headphone
Than I put in a 50K potmeter and saw that under 4.5K the oscillating starts, and with around 40K Ohm it is taking to long to get to zero on the outputs.
I settled with a fixed 22K resistor.
The relative and absolute offset is now few millivolts from cold to warm and settles in few seconds.
Thanks, mighty ZM
Than I put in a 50K potmeter and saw that under 4.5K the oscillating starts, and with around 40K Ohm it is taking to long to get to zero on the outputs.
I settled with a fixed 22K resistor.
The relative and absolute offset is now few millivolts from cold to warm and settles in few seconds.
Thanks, mighty ZM
Another weekend, more tinkering to figure out why my ACP+ is being weird.
Last night I totally reworked the board to chassis-mount RCA connections and installed the insulated power jack. This reduced the ground loop hum when connected to my ACA. Moving the ACA out from under my monitor eliminated it altogether…(I have a 6 month old at home and don’t always think all my options through.)
The problem that persists though is an extremely quiet, basically silent right channel on input 1. I resoldered the inputs and tested for continuity. I replaced the switch. The problem persisted. I put the chassis back together because it was bedtime.
Today at lunch, on a hunch I checked for continuity between the signal and ground on the right channel of input 1, and presto!! And electrical connection where there should not be. When I next have time, soldiering iron is coming back out and I’ll figure out if there is a bridge at the rca jack, a soldier joint, or some damage to the board from soldering and resoldering so many times. Update to follow!
Last night I totally reworked the board to chassis-mount RCA connections and installed the insulated power jack. This reduced the ground loop hum when connected to my ACA. Moving the ACA out from under my monitor eliminated it altogether…(I have a 6 month old at home and don’t always think all my options through.)
The problem that persists though is an extremely quiet, basically silent right channel on input 1. I resoldered the inputs and tested for continuity. I replaced the switch. The problem persisted. I put the chassis back together because it was bedtime.
Today at lunch, on a hunch I checked for continuity between the signal and ground on the right channel of input 1, and presto!! And electrical connection where there should not be. When I next have time, soldiering iron is coming back out and I’ll figure out if there is a bridge at the rca jack, a soldier joint, or some damage to the board from soldering and resoldering so many times. Update to follow!
Last edited:
Welp, having eliminated all other possibilities (solder bridges, bad RCA jack) it appears there is some sort of short from Input 1's right channel and ground in the board itself. I think I likely damaged it a year ago when I removed the board mounted RCA jacks in order to mount in a chassis. Since then intermittent dropping of that channel, and eventually the loss of it altogether has been an issue. My hypothesis would be that as I've messed with he board to try to fix it, I've likely flexed it just enough to make that short worse, and now, permanent. Luckily I am currently only using a single input on the ACP+, so the system still works! If I add another input I will wire up a switch off the board.
Chasing this down has certainly helped me learn a lot more about diagnosis and how this circuit works ! So altogether a positive experience!
Now I need to figure out to do with a matched quad of Toshiba 2SJ74. Any recommendation for an amp to power the uppers on my lxmimi's?
Chasing this down has certainly helped me learn a lot more about diagnosis and how this circuit works ! So altogether a positive experience!
Now I need to figure out to do with a matched quad of Toshiba 2SJ74. Any recommendation for an amp to power the uppers on my lxmimi's?
It is OK now, first put in a 2K2 resistor instead of the 150K --> 'oscillating' like hell....you should have heard the noise coming from my old headphone
Than I put in a 50K potmeter and saw that under 4.5K the oscillating starts, and with around 40K Ohm it is taking to long to get to zero on the outputs.
I settled with a fixed 22K resistor.
The relative and absolute offset is now few millivolts from cold to warm and settles in few seconds.
Thanks, mighty ZM
nice
now make proper presentation and share your efforts with Greedy Boyz
Hello all,
I plan to assemble an ACP+/ ACA mini combo, with the ACA mini PSU powering both units. As has been pointed out, the ACP+ kit comes with its own PSU and is not available without it. I started to compile the parts list for an ACP+ completion kit. However the instructions are that "Q1, Q2 are matched within 1 MA IDss". Is this practical or do I have to buy the kit and toss the PSU on the vast global pile of electronic trash? Thx.
I plan to assemble an ACP+/ ACA mini combo, with the ACA mini PSU powering both units. As has been pointed out, the ACP+ kit comes with its own PSU and is not available without it. I started to compile the parts list for an ACP+ completion kit. However the instructions are that "Q1, Q2 are matched within 1 MA IDss". Is this practical or do I have to buy the kit and toss the PSU on the vast global pile of electronic trash? Thx.
cccbbbp,
You could always buy the matched quad 2SJ74 in the store. I suspect the Toshiba's are even a little better than the Linear Systems the kit comes with (no clue if it actually matters much sound wise). I'm not sure which range is ideal 6-8mA or 8-11mA. Of course that still leaves you with needing to find a matched pair of J113. I suspect you could buy a pair, or quad, of these matched somewhere online.
I am also curious how to best match JFETs and was actually looking at various threads when you posted your question on this vary topic. I suspect it's not as hard as it seems at first glance and just needs some research into "how to diy" matching JFETs.
For the power supply, I've been looking at the Meanwell GST120A24-P1M which is probably way overkill for the purpose of ACA Mini + ACA+ , but I figure it's then functional for other projects at 5A capable output.
Best,
Dan
EDIT: I just looked at my ACA+ and it appears to have the Toshiba 2SJ74 in the "BR" rating, which if I'm not mistaken are the lower range 3-6mA, implying the 6-8mA would be the best choice with what's offered at the store. With all that said, I suspect the best bang for your buck is just buy the full kit once it's available (currently out of stock).
You could always buy the matched quad 2SJ74 in the store. I suspect the Toshiba's are even a little better than the Linear Systems the kit comes with (no clue if it actually matters much sound wise). I'm not sure which range is ideal 6-8mA or 8-11mA. Of course that still leaves you with needing to find a matched pair of J113. I suspect you could buy a pair, or quad, of these matched somewhere online.
I am also curious how to best match JFETs and was actually looking at various threads when you posted your question on this vary topic. I suspect it's not as hard as it seems at first glance and just needs some research into "how to diy" matching JFETs.
For the power supply, I've been looking at the Meanwell GST120A24-P1M which is probably way overkill for the purpose of ACA Mini + ACA+ , but I figure it's then functional for other projects at 5A capable output.
Best,
Dan
EDIT: I just looked at my ACA+ and it appears to have the Toshiba 2SJ74 in the "BR" rating, which if I'm not mistaken are the lower range 3-6mA, implying the 6-8mA would be the best choice with what's offered at the store. With all that said, I suspect the best bang for your buck is just buy the full kit once it's available (currently out of stock).
Last edited:
It appears my thought about using a 5A SMPS above would go against using the SMPS filter from the store, which was part of my planned ACA Mini + ACA+ combined build (just like cccbbbp).
I'm guessing running two of these 3A rated filters in parallel, one for each amp (ACA Mini and ACA+) does not prevent the chance of running too much current through a single SMPS filter due to the 5A supply capability of the power supply. Even though neither amp should draw more than 3A, I suspect limiting the power supply to 3A or less is the proper and safe approach.
Anyone have input on using a 5A supply for this purpose? Does running two SMPS filter boards in parallel really mitigate the risk in any way, or would that cause other issues a noob like myself isn't aware of?
Thanks in advance.
-Dan
I'm guessing running two of these 3A rated filters in parallel, one for each amp (ACA Mini and ACA+) does not prevent the chance of running too much current through a single SMPS filter due to the 5A supply capability of the power supply. Even though neither amp should draw more than 3A, I suspect limiting the power supply to 3A or less is the proper and safe approach.
Anyone have input on using a 5A supply for this purpose? Does running two SMPS filter boards in parallel really mitigate the risk in any way, or would that cause other issues a noob like myself isn't aware of?
Thanks in advance.
-Dan
Thanks Jim, that was what I was thinking initially and just wasn't sure if there was a chance of some sort of inrush current upon power up may cause a spike beyond the 3A limit specified for the filter board components. The filter board may in fact be tolerant to a very brief spike above 3A anyway.
Does having two filters in parallel for PSU input help in any way if each amp is individually fed by each filter board output? This question also assumes there was a concern of a power spike in initial power up current. I see some folks have put two filters in series to further smooth out the PSU, and just didn't see any info looking through the threads of using two in parallel for powering two boards from a single supply.
Best,
Dan
P.S. You rock! I really appreciate the quick feedback.
Does having two filters in parallel for PSU input help in any way if each amp is individually fed by each filter board output? This question also assumes there was a concern of a power spike in initial power up current. I see some folks have put two filters in series to further smooth out the PSU, and just didn't see any info looking through the threads of using two in parallel for powering two boards from a single supply.
Best,
Dan
P.S. You rock! I really appreciate the quick feedback.
Inrush is draw… which will almost certainly be above 3A. The ACAmini is obviously going to draw more than the ACP+, but yes, having a filter for the ACP and the ACA will help spread the load, so to speak.
People have used these filters with ACAmini to great success, so don't worry about it.
People have used these filters with ACAmini to great success, so don't worry about it.
I'm using a filter with my amp camp mini even though it's also using the 5 amp power supply from the original amp camp amp. No issues. It's been fine since it was assembled about 18 months ago. Tried the filter with that power supply because the amp camp mini doesn't draw more than 3 amps.
Have not tried it on the original amp camp amp as evidently that uses a bit more than 3A. Evidently using one filter per channel is the way to go with the ACA.
In absolutely subjective listening the 5a supply sounds better than the one that ships with the mini kit and both power supplies sound better with the filter. It's a great sounding little amp.
Have not tried it on the original amp camp amp as evidently that uses a bit more than 3A. Evidently using one filter per channel is the way to go with the ACA.
In absolutely subjective listening the 5a supply sounds better than the one that ships with the mini kit and both power supplies sound better with the filter. It's a great sounding little amp.
Hi Dan,
Thanks for pointing me to the JFET section in the store. Nelson has info on the appropriate JFETs for the various amps there and info on matching.
Several more questions:
1. For the power switches, I plan on using panel-mount switches with soldered leads ideally to connectors which would mate to board-mount connectors on the 2 boards. Does anyone know of such a connector pair with a board-mount connector that fits the hole patterns on the ACP+ and
the mini?
2. Would there be a similar connector pair for panel-mount LEDs?
3. Also, why SPDT switches (on/off/on) specified for the 2 amps? Wouldnt SPST suffice?
Cccbbbp,
First, I'm pretty much a total noob on diyaudio. I have some basic understanding about some of this awesome stuff folks do here, but pretty basic and not based on empirical experience, so other veteran fearless amp builders are the ones to listen to. I'll see if my response se to your questions helps a little.
1. You could use a switch that uses a spade style connection and maybe add some conductive grease to prevent long term oxidation. Why not solder the other end of the wires directly to the board if the intent is to be able to dismantle without desoldering. Is there a reason you want the wires connecting switch to board to disconnect at both ends? Two soldered connections seems more electrically "secure" and not all that hard to desolder if it came to that.
2. For LEDs, I'm curious why not just have those soldered on both ends (no spades or terminal blocks) and make the LED mechanically attachable/detachable from the panel front. If you want to be able to electrically disconnect the LED from the board I suspect there's a multitude of options for various size/shaped spades.
3. I think you are technically correct that all you need is a single pole single throw, however I suspect there's a good reason for the double throw. Positive and negative power each get their own on-off with that double throw so I'd only be speculating as to why that helps (same wire length has to have some benefits to minimizing noise maybe?).
Anyway, by answering all these questions incorrectly I'm sure someone far more knowledgeable will hop on and rectify my well intentioned misinformation.
Best,
Dan
First, I'm pretty much a total noob on diyaudio. I have some basic understanding about some of this awesome stuff folks do here, but pretty basic and not based on empirical experience, so other veteran fearless amp builders are the ones to listen to. I'll see if my response se to your questions helps a little.
1. You could use a switch that uses a spade style connection and maybe add some conductive grease to prevent long term oxidation. Why not solder the other end of the wires directly to the board if the intent is to be able to dismantle without desoldering. Is there a reason you want the wires connecting switch to board to disconnect at both ends? Two soldered connections seems more electrically "secure" and not all that hard to desolder if it came to that.
2. For LEDs, I'm curious why not just have those soldered on both ends (no spades or terminal blocks) and make the LED mechanically attachable/detachable from the panel front. If you want to be able to electrically disconnect the LED from the board I suspect there's a multitude of options for various size/shaped spades.
3. I think you are technically correct that all you need is a single pole single throw, however I suspect there's a good reason for the double throw. Positive and negative power each get their own on-off with that double throw so I'd only be speculating as to why that helps (same wire length has to have some benefits to minimizing noise maybe?).
Anyway, by answering all these questions incorrectly I'm sure someone far more knowledgeable will hop on and rectify my well intentioned misinformation.
Best,
Dan