Can anyone tell me if they have built a GainClone amp using these boards:
PCB for Audio Amplifier LM3875 HiFi Gainclone Amp Qty:2 on eBay (end time 24-Feb-10 00:05:21 GMT)
There are several options with these boards, inverted and non inverted and plenty of places for components. Just wanted to know if anyone has used these and what the sound quality would be like, as it doesn't seem like there is a separate star ground for the power and the signal.
Thanks
PCB for Audio Amplifier LM3875 HiFi Gainclone Amp Qty:2 on eBay (end time 24-Feb-10 00:05:21 GMT)
There are several options with these boards, inverted and non inverted and plenty of places for components. Just wanted to know if anyone has used these and what the sound quality would be like, as it doesn't seem like there is a separate star ground for the power and the signal.
Thanks
It has a passable star-ground, though not the best layout possible. The bigger concern is the absence of a lifted-ground (ground-loop breaker), though one can perhaps be kludged in by lifting a pin or two and soldering a 10-ohm resistor suitably. Is this also available on EBay US, or only on EBay UK?
Thanks for the quick reply linuxguru.
The PCBs were advertised on ebay uk, but they are from the US.
What do you mean when you say lifted-ground?
The PCBs were advertised on ebay uk, but they are from the US.
What do you mean when you say lifted-ground?
Lifted-ground is a separate signal ground, isolated from the ground plane with a small resistor (say 10 ohms). The shield ground from the input connector is connected to the lifted ground, and so is the input bias resistor and feedback network, but nothing else.
It's also known as a ground-loop breaker, and it's main purpose is to eliminate 50/60 Hz hum pickup.
I have another question.
My end goal of my audio project will be a 4 separate aluminium enclosures. 2 of these will be Gainclone amps (bi-amping my speakers), 1 will be a power supply for the Gainclone amps and the last one will be a pre amp which will be IR controlled and use a PIC microprocessor and relays to carry out the switching etc.
The pre-amp will have it's own small power supply built into it fed via 230v, transformed done to 5V to run the microprocessor, the enclosure will obviously be earthed as will the Gainclone power supply enclosure.
The 4 enclosures will be stacked, so technically all of the enclosures will be at the same potential, but I will be earthing the Gainclone enclosures with an extra cable from the power supply.
Now, you say that I need to ground the signal ground via a 10 ohm resistor, where should I put this resistor? Should I put 1 in each enclosure?
Any help would be great.
Thanks
My end goal of my audio project will be a 4 separate aluminium enclosures. 2 of these will be Gainclone amps (bi-amping my speakers), 1 will be a power supply for the Gainclone amps and the last one will be a pre amp which will be IR controlled and use a PIC microprocessor and relays to carry out the switching etc.
The pre-amp will have it's own small power supply built into it fed via 230v, transformed done to 5V to run the microprocessor, the enclosure will obviously be earthed as will the Gainclone power supply enclosure.
The 4 enclosures will be stacked, so technically all of the enclosures will be at the same potential, but I will be earthing the Gainclone enclosures with an extra cable from the power supply.
Now, you say that I need to ground the signal ground via a 10 ohm resistor, where should I put this resistor? Should I put 1 in each enclosure?
Any help would be great.
Thanks
Just another quick one...
If I were to tie the signal grounds to the chassis in each gainclone amp in my system, then there would be a parallel ground (2 x 10 ohm in series), so the start ground of the signal would be 5 ohms difference from main ground? And again, if I were to add a 10 ohm resistor to the pre-amp, I would have an even lesser resistance as there would technically be 3 x 10 ohm resistors in parallel!
Also, do I ground the power supply (+PGND and -PGND) in the power supply enclosure, the Gainclone amp enclosures, or both.
Thanks again for all of your help
Craig
If I were to tie the signal grounds to the chassis in each gainclone amp in my system, then there would be a parallel ground (2 x 10 ohm in series), so the start ground of the signal would be 5 ohms difference from main ground? And again, if I were to add a 10 ohm resistor to the pre-amp, I would have an even lesser resistance as there would technically be 3 x 10 ohm resistors in parallel!
Also, do I ground the power supply (+PGND and -PGND) in the power supply enclosure, the Gainclone amp enclosures, or both.
Thanks again for all of your help
Craig
When you tie the signal ground to the chassis ground on each chassis, there is a potential for ground loops to be created. That's because there may be two separate shields in the signal cable between the source and the amplifier, as well as a possible common chassis ground (through earthing) between the source and the two amplifiers. Ground loops are the most common source of 50/60 Hz hum - they can inductively pick up line frequency interference from transformers, AC lines, etc. It's pretty difficult to eliminate all sources of such interference.
It's easier to break ground loops at the input to the amplifier. The most common way of doing this is to separate the shield ground from the chassis ground (inside the gainclone enclosure) with a 10-ohm resistor, paralleled with two back-to-back rectifier diodes like the 1N4002. When this happens, there is no longer a low-impedance ground loop, which helps reduce hum substantially.
The exact value of the resistance isn't that important - anywhere from 2.2 to 10 ohms is commonly used.
Most probably better to ground them together in each gainclone amp enclosure at the star ground point only. This should be close to, or at, the terminal of one of the filter capacitors on the gainclone board.
When you tie the signal ground to the chassis ground on each chassis, there is a potential for ground loops to be created. That's because there may be two separate shields in the signal cable between the source and the amplifier, as well as a possible common chassis ground (through earthing) between the source and the two amplifiers. Ground loops are the most common source of 50/60 Hz hum - they can inductively pick up line frequency interference from transformers, AC lines, etc. It's pretty difficult to eliminate all sources of such interference.
It's easier to break ground loops at the input to the amplifier. The most common way of doing this is to separate the shield ground from the chassis ground (inside the gainclone enclosure) with a 10-ohm resistor, paralleled with two back-to-back rectifier diodes like the 1N4002. When this happens, there is no longer a low-impedance ground loop, which helps reduce hum substantially.
The exact value of the resistance isn't that important - anywhere from 2.2 to 10 ohms is commonly used.
Most probably better to ground them together in each gainclone amp enclosure at the star ground point only. This should be close to, or at, the terminal of one of the filter capacitors on the gainclone board.
Thanks very much for you answer.
So basically, every enclosure will be earthed (mains earth, for protection) and in each of the 2 Gainclone enclosures I will have a star-ground point, where the shield ground and +PGND and -PGND will be connected together and then connected to the chassis via: "a 10-ohm resistor, paralleled with two back-to-back rectifier diodes like the 1N4002"
Is that correct?
Thanks again for your help, I have been an electrical engineer for 12 years now and this ground-loop thing is all new to me.
So basically, every enclosure will be earthed (mains earth, for protection) and in each of the 2 Gainclone enclosures I will have a star-ground point, where the shield ground and +PGND and -PGND will be connected together and then connected to the chassis via: "a 10-ohm resistor, paralleled with two back-to-back rectifier diodes like the 1N4002"
Is that correct?
Thanks again for your help, I have been an electrical engineer for 12 years now and this ground-loop thing is all new to me.
a 10r resistor paralleled with a diode and then paralleled with a second diode that is facing the other way.
This allows currents to bypass the resistor in both directions with a voltage drop limited to ~1Vpk.
I have called this diodes in inverse parallel.
I like neither the PCB layout nor the schematics of that ebay item.
Both the inverting and the non-inverting have errors and both omit some band-limiting components and show little to help stabilise the amplifier.
The PCB is a poor implementation of the schematics.
It is very much easier to build a monoblock and learn how to ground and connect that.
Then try to match the monoblock performance using two channels from the same power supply. This is not an easy task.
Both the inverting and the non-inverting have errors and both omit some band-limiting components and show little to help stabilise the amplifier.
The PCB is a poor implementation of the schematics.
It is very much easier to build a monoblock and learn how to ground and connect that.
Then try to match the monoblock performance using two channels from the same power supply. This is not an easy task.
Last edited:
I posted the erroneous "back-to-back" terminology - mea culpa. It's incorrect, because it suggests a series connection of diodes, akin to two electrolytic capacitors being used as a single non-polar capacitor. With diodes, that has no particular use in this context (of lifted grounds).
The +PGND and -PGND are connected to the PCB ground at or near the star ground point. These carry high currents, so it helps to keep them close to the star ground on the PCB. The lifted ground can be connected to the PCB ground (through the ground-loop breaker) anywhere on the PCB, since it carries relatively low currents in normal operation (of course, it helps in practice to ground the ground-loop breaker also near the star ground, to eliminate possible ground-bounce effects).
The shield ground on the input cable is connected to the lifted ground through the input terminal block. This does not have to be anywhere near the star ground on the PCB.
The chassis ground on the gainclone box may be left isolated from the PCB ground. If you choose to connect it to the PCB ground, ensure that it is connected at a single point only (to avoid yet another ground loop). This is usually at or close to the star ground on the PCB, but it's not mandatory, since it only carries stray small currents.
Hi again,
I was looking through the threads about grounding and found this page:
http://www.diyaudio.com/forums/chip-amps/115698-understanding-star-grounding-4.html
The 2nd diagram (by fluckscapacitor) looks similar to what you are mentioning in this thread and I notice that you (AndrewT) comment on the design. Now I would like to know, if my power supply consists of two transformers and rectifier boards and each rectifier board will feed a separate Gainclone amp enclosure (each amp enclosure will have 2 x LM3875 chips, for bi-amping my speakers) do I put a separate ground-break circuit after each rectifier board?
I have also noticed that Peter Daniel (Audiosector) amps take both the +PGND and -PGND separately to the amp and then connect them there, not at the PSU! Is there any advantage by doing this?
I was looking through the threads about grounding and found this page:
http://www.diyaudio.com/forums/chip-amps/115698-understanding-star-grounding-4.html
The 2nd diagram (by fluckscapacitor) looks similar to what you are mentioning in this thread and I notice that you (AndrewT) comment on the design. Now I would like to know, if my power supply consists of two transformers and rectifier boards and each rectifier board will feed a separate Gainclone amp enclosure (each amp enclosure will have 2 x LM3875 chips, for bi-amping my speakers) do I put a separate ground-break circuit after each rectifier board?
I have also noticed that Peter Daniel (Audiosector) amps take both the +PGND and -PGND separately to the amp and then connect them there, not at the PSU! Is there any advantage by doing this?
No I haven't built these boards.
Are you thinking of buying these PCBs?
They seem cheap to me and a good way to get your first chip amp under your belt.
The schematics are from National Semiconductor's datasheets. Similar schematics can be found here:
http//www.national.com/ds/LM/LM3875.pdf
http://www.national.com/ds/LM/LM4780.pdf
Ask the seller if they have been thoroughly tested.
regards
Hi AndrewT, I bought 2 of these boards just to experiment with, but will end up making my own PCBs I expect when I build all of my components for my system.
this is worth some experimentation.
Try bolting all the grounds together as the star ground (audio ground).
Each ground wire can have a crimp tag.
Try one disconnecting network. Is it better than using a direct wire link.
Try two disconnecting networks (one per channel). Is this better than either of the preceding options?
- Status
- Not open for further replies.