As it should. I just wanted to be sure with you mentioning reversed do-dahs and upside down FET's 🙂
I really don't know what to suggest beyond you checking in detail the connectivity of each pin and part and checking they match the good channel electrically. From a purely theoretical point based on what results you are getting Q3 fitted incorrectly would give the same kind of results that you are seeing, The voltage on the gate of the lower FET would be input FET and preset position dependent but that doesn't change what is happening here.
I'll show you this for interest but remember that in simulation models do not behave exactly like real ones, particularly if incorrectly fitted and biased... however... here is Q3 fitted so that it presents as a forward biased diode from (effectively) gate to source.
I really don't know what to suggest beyond you checking in detail the connectivity of each pin and part and checking they match the good channel electrically. From a purely theoretical point based on what results you are getting Q3 fitted incorrectly would give the same kind of results that you are seeing, The voltage on the gate of the lower FET would be input FET and preset position dependent but that doesn't change what is happening here.
I'll show you this for interest but remember that in simulation models do not behave exactly like real ones, particularly if incorrectly fitted and biased... however... here is Q3 fitted so that it presents as a forward biased diode from (effectively) gate to source.
I'll look in again tomorrow. Once you have identified in your mind which FET is which just do simple resistance checks from the pins to the parts that are supposed to connect to those pins. That should lead you to Q3.
Check that the same pin position on Q3 connects to the same part as the good channel.
If you can not find any obvious issue then swap Q3 but be sure to check that the pins of Q3 really do go where they should. Use a data sheet diagram for the pin outs for whatever transistor you decide to fit rather than just relying on the board markings.
Check that the same pin position on Q3 connects to the same part as the good channel.
If you can not find any obvious issue then swap Q3 but be sure to check that the pins of Q3 really do go where they should. Use a data sheet diagram for the pin outs for whatever transistor you decide to fit rather than just relying on the board markings.
IT WORKS!
However, I have a little egg on my face 😡.
As I was checking continuity looking at the LT Spice layout It occurred to me that I had not thought about an additional resistor wired under the board between E of the ZTX and R8 (1k)
My mate bought me the first set of boards for christmas (which were not handed) and informed of a tweak that would give an extra watt. Adding a 2.2k resistor as mentioned above. As you know those boards have been working fine,
When I was checking I measured continuity and ohms between E (or what I thought was E of the ZTX and the OR68 resistor that the 2.2k is attached to. The Good and Bad board were showing different readings. I suspected that the leg of the 2.2k was not soldered to E of ZTX. I removed it and resoldered to the other outside leg and BINGO. Looking closely at the solder masks of both boards showed that they are different and despite the screen print suggesting that the ZTX on the bad board be installed in the same orientation that seems not to be the case.
I will give it a proper listen to later, but early indications are that it is fine. No hum and whisper quiet.
Thanks to everyone that helped me out with it is very much appreciated. I suspect that if we were not having a heatwave here in the UK and hanging around the house a lot I may not have been so didligent in my fault finding.
PS Does anyone put large washers on the front face of Q1 and 2 to disspate heat? I did this on the other amp.
However, I have a little egg on my face 😡.
As I was checking continuity looking at the LT Spice layout It occurred to me that I had not thought about an additional resistor wired under the board between E of the ZTX and R8 (1k)
My mate bought me the first set of boards for christmas (which were not handed) and informed of a tweak that would give an extra watt. Adding a 2.2k resistor as mentioned above. As you know those boards have been working fine,
When I was checking I measured continuity and ohms between E (or what I thought was E of the ZTX and the OR68 resistor that the 2.2k is attached to. The Good and Bad board were showing different readings. I suspected that the leg of the 2.2k was not soldered to E of ZTX. I removed it and resoldered to the other outside leg and BINGO. Looking closely at the solder masks of both boards showed that they are different and despite the screen print suggesting that the ZTX on the bad board be installed in the same orientation that seems not to be the case.
I will give it a proper listen to later, but early indications are that it is fine. No hum and whisper quiet.
Thanks to everyone that helped me out with it is very much appreciated. I suspect that if we were not having a heatwave here in the UK and hanging around the house a lot I may not have been so didligent in my fault finding.
PS Does anyone put large washers on the front face of Q1 and 2 to disspate heat? I did this on the other amp.
I could be wrong, but I think the purpose of the 'bumper washers' is to apply uniform pressure to the mosfets (against the heatsink) rather than to dissipate heat themselves. A good idea in itself which also makes over-tightening the bolts less likely. (ZenMod has referred to poor quality bolts fracturing while being torqued down - I find it hard to imagine applying that much pressure but you do want max possible heat transfer.)
Well that's good 🙂
Its difficult to visualise just what has happened then with the layout differences. I would say it is worth you checking the voltage dropped across the 0.68 and 0.47 ohm resistors is similar to the other channel. I would expect around 0.54 and .38 volts across each of the 0.68 and 0.47 ohm respectively.
It will be as davidjt suggest. The data sheet for the transistors actually gives a mounting torque and it will probably not be as high as you might think.
Thanks Mooly. I will take some readings. I have just had my first 30 mins with it and these things amaze me. IMO the SQ is up there with some of the best.....mind you my ears are not as good as they once were 🙂
Yes 1Nm is nothing at all really. I do a lot of cycling and my own maintenance and some of these settings are only 7Nm which seems too little to me.
I have fitted an XLR and all the wiring to use it in bridged mode. Once I have done the previous amp I can see if it has been worthwhile.
Yes 1Nm is nothing at all really. I do a lot of cycling and my own maintenance and some of these settings are only 7Nm which seems too little to me.
I have fitted an XLR and all the wiring to use it in bridged mode. Once I have done the previous amp I can see if it has been worthwhile.
Voltage Drops.
Good Board 0.68
0.48v (consistent across all)
Good Board 0.47
0.33 (Consistent)
Bad Board 0.68
0.48v (Consistent)
Bad Board 0.47
0.33v (Consistent)
Those are all fine and show the bias regulator is working correctly. So each channel is passing just over 1.4A
Q3's characteristics to some extent determine the exact values (the base/emitter voltage needed to bring Q3 into conduction).
So all good 🙂
Q3's characteristics to some extent determine the exact values (the base/emitter voltage needed to bring Q3 into conduction).
So all good 🙂
It's probably answered many times but anyway, what's so special about feeding the load from the (almost) center point of the resistors arrangement and not from the drain of the lower transistor?
upper part of OS is Mu Follower
it works as impedance xformer, which lower mosfet is seeing
it can be explained in few other ways, but this one should suffice for now
if you move output node down, Mu Follower becomes simple CCS
it works as impedance xformer, which lower mosfet is seeing
it can be explained in few other ways, but this one should suffice for now
if you move output node down, Mu Follower becomes simple CCS
I am sure that some on here have confrigured the ACA as a monoblock. I have added an XLR socket to one of my ACAs and added the buss and xlr wiring. Should the output be lower than as a stereo amp? I would have assumed that as it is intended to raise the output that it would have sounded louder unless of course it alters the gain?
EDIT: I have also tried the "option for advanced builders" the RCA lead with 2 x 34k resistors used to link one input to the opposite channel -out speaker socket. This is louder than the bridged XLR, but still not quite as loud as stereo
EDIT: I have also tried the "option for advanced builders" the RCA lead with 2 x 34k resistors used to link one input to the opposite channel -out speaker socket. This is louder than the bridged XLR, but still not quite as loud as stereo
Last edited:
Hi Puffin,
First to use the XLR bridged setup you have to have a 'balanced output' source. Do you have this?
Second you refer to "option for advanced builders", I assume this is to do with the kits you have, unfortunately we cannot see this. We have to use the official ACA build guide. (https://guides.diyaudio.com/Guide/Amp+Camp+Amp+V1.6+Build+Guide/5) Have you looked there for the RCA bridged mode?
How are you connecting your speakers and what impedance are they?
First to use the XLR bridged setup you have to have a 'balanced output' source. Do you have this?
Second you refer to "option for advanced builders", I assume this is to do with the kits you have, unfortunately we cannot see this. We have to use the official ACA build guide. (https://guides.diyaudio.com/Guide/Amp+Camp+Amp+V1.6+Build+Guide/5) Have you looked there for the RCA bridged mode?
How are you connecting your speakers and what impedance are they?
The option for advanced builders is Step 46 on the version of the ACA Build guide I have and designated V1.5 Build Guide. It is a Pass reccomended option. Link below.
https://guides.diyaudio.com/Guide/Amp+Camp+Amp+V1.5+Build+Guide/2
I have a balnaced output source and used the reccomended speaker wiring scheme. If you are using the optional 34k RCA lead you do not need a balanced source.
https://guides.diyaudio.com/Guide/Amp+Camp+Amp+V1.5+Build+Guide/2
I have a balnaced output source and used the reccomended speaker wiring scheme. If you are using the optional 34k RCA lead you do not need a balanced source.
I always get a bit confused by all the options tbh 🙂
Lets take a normal stereo 2 channel build. It is what it is and plays at a given level for a given input voltage.
A parallel monobloc would feed that same input signal into both channels. You then parallel the speaker outputs and get exactly the same voltage output but now with twice the current ability for more difficult loads. So it should sound as loud as in the normal stereo set up.
A bridged set up is designed to deliver twice the voltage swing to the load and this is done by either feeding each input with out of phase signals (180 degree difference) or in the case of the ACA you can also feed the input signal to just one amplifier channel input and then feed the output of that channel (via a resistor) into the input of the other channel. It won't sound twice as loud but should be noticeably louder before it clips than the standard stereo amp.
The fact the ACA is inverting by design does the phase shifting for you. These set ups double the voltage swing to the load because the load is fed from between the two outputs and not from an output to ground. The current delivery remains as for the normal stereo setup and so this is not a good choice for difficult loads.
The first bridged option is the best (imo) where you feed the inputs from a front end phase splitter because the second method of feeding one channels input from the other channels output has the real possibility of that signal being modulated by the high output impedance of the ACA as it delivers current into the load.
Hi Mooly, thanks for your response. My only experience of creating bridged amplifiers is that years ago I made a bridged Gainclone (LM3886) by following the TI Overture schematic and building it point to point on perfboard. Was amazed when it worked and I also converted two Cambridge P500 power amps by adding bridging modules.
I cannot say now that the above were in fact louder than the original amps as I was not able to do a side by side comparison as I can with the ACA. I was just surprised that when using both the XLR and the RCA feeding the signal to the other channel that the sound level was less.
I can always bi-amp my speakers 🙂
Rob.
I cannot say now that the above were in fact louder than the original amps as I was not able to do a side by side comparison as I can with the ACA. I was just surprised that when using both the XLR and the RCA feeding the signal to the other channel that the sound level was less.
I can always bi-amp my speakers 🙂
Rob.