I have just had a conversation with a friend about Gainclones, DIY audio etc. He told me that a friend of his had built a volume control that used relays and a binary counting IC. This got me thinking that it might be a good idea to use this as my Gainclone volume control. 8 resistors wired in series, each having a pair of relay contacts connected across it. The relays can be controlled by a binary counter IC, this being stepped by either a simple UP or DOWN switch arrangement or, as done in his design, a potentiometer feeding an A to D converter. Anyway what I am wondering is that is this arrangement a good idea as far as audio quality is concerned ? How would this "variable resistor" be best used, connected to the input with a fixed resistor to ground or with a fixed resistor from the input and have the resistor/relay chain going to ground ?
I hope I have explained this idea clearly enough for someone to give advice if this is worth building.
I hope I have explained this idea clearly enough for someone to give advice if this is worth building.
I imagine that you are aware of the Remote Relay Volume control:
http://www.diyaudio.com/forums/showthread.php?threadid=12186&goto=newpost
The general consensus is that relay switched resistors sound good.
If your very simple approach works I imagine a lot of people would like it as a simple and cheap option. Can you get over 8 levels with this idea? The pot on the gainclone cannot be eliminated I have heard, unless a resistor is substituted, so your resisitor values will have to be chosen to be similar to the pot. Of course these would be the most useful values anyway.
Also look at the Passive Preamp thread for info on shunt type controls, etc.
http://www.diyaudio.com/forums/showthread.php?threadid=12186&goto=newpost
The general consensus is that relay switched resistors sound good.
If your very simple approach works I imagine a lot of people would like it as a simple and cheap option. Can you get over 8 levels with this idea? The pot on the gainclone cannot be eliminated I have heard, unless a resistor is substituted, so your resisitor values will have to be chosen to be similar to the pot. Of course these would be the most useful values anyway.
Also look at the Passive Preamp thread for info on shunt type controls, etc.
With 8 resistors and 8 relays, 255 volume levels can be selected. I`m going to look into this idea and design a circuit, i`ll post the results soon. As long as having 8 resistors with relays (very small ones) connected across them doesn`t reduce sound quality, then this could be a really good volume contol for a Gainclone.
Prof
The first issue will be to select a relay and actually listen to it in an audio chain. You may like to compare to the sound of a solder joint and high quality silver switch. I somehow think you'll go through quite a few relays before you find one not too obtrusive. Your choice of 8 relays assumes you can get resistor values which relate as powers of two; no idea how easy/difficult this may be. Then you'll need a controller which won't do simple up/down counting but convert to log, right? For any given volume you'll be listening to a few of the contacts and a few of the resistors - not a very minimalistic approach. I would probably prefer 30 relays and 30 sets of fixed resistor dividers as 30 is even more steps than i need.
cheers
peter
The first issue will be to select a relay and actually listen to it in an audio chain. You may like to compare to the sound of a solder joint and high quality silver switch. I somehow think you'll go through quite a few relays before you find one not too obtrusive. Your choice of 8 relays assumes you can get resistor values which relate as powers of two; no idea how easy/difficult this may be. Then you'll need a controller which won't do simple up/down counting but convert to log, right? For any given volume you'll be listening to a few of the contacts and a few of the resistors - not a very minimalistic approach. I would probably prefer 30 relays and 30 sets of fixed resistor dividers as 30 is even more steps than i need.
cheers
peter
This kind of regulation (lets assume they have been borrowed from Alephs P designs from Mr. Pass) has advantages that with only 8 switches (high quality,silver,gold plated contacs are recommended) and 8 resistors one gets 256 combinations.Disadvantages of this design is that unlike in stepped atenuators it works partly as shunt regulator and partly as ladder one.It makes many series/parallel combinations,so the preamp or DAC (depending what sits before that regulation) doesn't see steady impedance.
In stepped attenuator this imput impedance is steady (not in shunt type).But when we consider stepped atenuator we have to know what kind of are we talking about.There are series ones ,ladder type and shunt type (I don't think that one is the way volume control should work so don't consider this one).
The series one for 24 positions needs:
2 deck ,24 position switch,and 48 resistors for unbalanced,
4 deck ,24 position switch and 96 resistors for balanced
But when we listen for example on second volume level signal goes through 22 resistors and at least the same solder joints.
In this case I would prefer the solution with relays and 256 positions becouse it probably doesn't sound worse.
The true ladder type where in every position there are two resistors
ne in signal path and the other to ground is a truly outstanding solution,becouse in every position signal goes through one (always diffrent resistor).But now for ladder type we need (24 position):
4 deck,24 position switch , 96 resistors for unbalanced
8 deck,24 position switch , 192 resistors for balanced
Also it's almost imposible to find values of resistors calculated for that kind of attenuator.This one is trully minimalist design and offers outstanding performance (I think only transformer volume control might be better),but we probably would have to order custom designs resistors.
Assuming above I thing the solution with relays and controling them chips is worth considering.When people say stepped attenuator,they usually mean the simplest one-series attenuator where in worst case we are listening to 23 resistors and solder joints.In that case I thing 8 relay type is going to be better (more minimalist),becouse here in worst case we have 8 resistors/solder joints ,and not all of them are series but part is parallel.
Bartek
In stepped attenuator this imput impedance is steady (not in shunt type).But when we consider stepped atenuator we have to know what kind of are we talking about.There are series ones ,ladder type and shunt type (I don't think that one is the way volume control should work so don't consider this one).
The series one for 24 positions needs:
2 deck ,24 position switch,and 48 resistors for unbalanced,
4 deck ,24 position switch and 96 resistors for balanced
But when we listen for example on second volume level signal goes through 22 resistors and at least the same solder joints.
In this case I would prefer the solution with relays and 256 positions becouse it probably doesn't sound worse.
The true ladder type where in every position there are two resistors
ne in signal path and the other to ground is a truly outstanding solution,becouse in every position signal goes through one (always diffrent resistor).But now for ladder type we need (24 position):4 deck,24 position switch , 96 resistors for unbalanced
8 deck,24 position switch , 192 resistors for balanced
Also it's almost imposible to find values of resistors calculated for that kind of attenuator.This one is trully minimalist design and offers outstanding performance (I think only transformer volume control might be better),but we probably would have to order custom designs resistors.
Assuming above I thing the solution with relays and controling them chips is worth considering.When people say stepped attenuator,they usually mean the simplest one-series attenuator where in worst case we are listening to 23 resistors and solder joints.In that case I thing 8 relay type is going to be better (more minimalist),becouse here in worst case we have 8 resistors/solder joints ,and not all of them are series but part is parallel.
Bartek
Check this post...
http://www.diyaudio.com/forums/showthread.php?postid=97685#post97685
Didn't occur to me until recently that this scheme could work well with an inverting GC. The first op-amp could be omitted if your preamp or source has a nice low output impedance...
http://www.diyaudio.com/forums/showthread.php?postid=97685#post97685
Didn't occur to me until recently that this scheme could work well with an inverting GC. The first op-amp could be omitted if your preamp or source has a nice low output impedance...
Attachments
Boy there are a lot of links about this subject, and I'm too ignorant to sort them out.
Heres a semi quote from the Jeff Rowland Concentra thread:
"the Concentra uses the CS3310 as volume control."
This is the chip vol control w/ precision resistors built in, which seems to work pretty darned well on the Concentra.
BUT a lot of people around here wouldn't trust such a thing.
BUT what if it were used to switch as a shunt device- then the signal isn't directly passing through it . Maybe that's how JR uses it? ... Could a simple circuit be worked out for this?
Heres a semi quote from the Jeff Rowland Concentra thread:
"the Concentra uses the CS3310 as volume control."
This is the chip vol control w/ precision resistors built in, which seems to work pretty darned well on the Concentra.
BUT a lot of people around here wouldn't trust such a thing.
BUT what if it were used to switch as a shunt device- then the signal isn't directly passing through it . Maybe that's how JR uses it? ... Could a simple circuit be worked out for this?
Hello Variac, I used an 11 step rotary switch to build an 11 step shunt attenuator. This has now been in use for a couple of days and gives me ample control of volume. My original idea of the 255 step, relay controlled attenuator would work, but would be more expensive/complicated to build. Now I`ve built the simple rotary attenuator, I`m quite happy to stick with what I`ve got.
Hi Jan,janneman said:
I've not spent a great deal of time on this, but I can't see why it shouldn't work just fine. Although perhaps you won't be able to get nice 1dB steps - this is my only real question at this stage. I'm guessing that this is what your reference to log tables means?
As a quick "proof of concept", I ran it through Excel (no access to simulators at work!) - here's what came out for a simple 4-bit sequence, using binary resistor values of 10, 20, 40 and 80K - I made the feedback resistor 220K, as this is a typical GainClone value. The 15 steps vary the gain from 32.3dB down to 8.8dB's - a typical GC with 10K/220K has 27dB
The purple line is simply what you'd get from 1dB steps and the blue line shows the calculated response. As you can see, the response is quite nice - there's plenty of resolution at higher settings, and larger step sizes the quieter end - just what you need, IME. The model was for 3 relays initially, and the curve was the same shape. If there is enough interest, I could try and expand it to include more relays tomorrow (if I get any spare time). This would bring more range, which is probably preferable for a real implementation.
Hope this helps,
Mark
Attachments
The amp with 32dB will be louderPeter Daniel said:
Seriously, how will you tell? If you want to properly compare two amps with different gain, you will have to make up that gain with a preamp (low-gain amp), or lose some gain with an attenuator (high-gain amp). Sure, I accept that there might well be some differences, but with a GC, both these options are likely to affect the sound more than the difference in gain.
I'd be interested to read more - have you got a URL or some search tips?
Cheers,
Mark
PS - Congrats on your venture and first review
The easiest way would be to compare it using a buffered, good quality volume control. I believe there will be a substantial difference, as comparing 250k and 300k resistors, inserted in a feedbak loop, was easily detectable. It is also dependant on what level of performance we expect from our GC. If it's only for casual listening or computer based system, then it's not worth investigating.
Here's the post by Joe discussing different gains and possible consequences:
http://www.diyaudio.com/forums/showthread.php?s=&postid=164564#post164564
Here's the post by Joe discussing different gains and possible consequences:
http://www.diyaudio.com/forums/showthread.php?s=&postid=164564#post164564
Hi Mark,
Yes I was referring to the error you get when trying to get linear-in-dB range with binary steps. But your curves are not that bad. What about using one extra relay to switch the 22ok between two values? You could use the 18-32dB range of the 4 bit thing, which is pretty linear, and use the extra relay to say shunt the 220k with a resistor to lower the gain by 18dB, bringing your linear range to full 32dB. Does that make sense? It would mean you need some logic because you no longer have a nice binary count. Or maybe you do have if the extra step is made 16dB.
Jan Didden.
Yes I was referring to the error you get when trying to get linear-in-dB range with binary steps. But your curves are not that bad. What about using one extra relay to switch the 22ok between two values? You could use the 18-32dB range of the 4 bit thing, which is pretty linear, and use the extra relay to say shunt the 220k with a resistor to lower the gain by 18dB, bringing your linear range to full 32dB. Does that make sense? It would mean you need some logic because you no longer have a nice binary count. Or maybe you do have if the extra step is made 16dB.
Jan Didden.
Mark H,
I like the simplicity of your calcs. After all, if we want "everything plus the kitchen sink" we should just get Apox's.
Also, the minimalist approach seems the right way to go with a chip amp. As Peter said, even a few positions can be quite useful.
I'd really appreciate a schematic that is totally clear, very simple, with parts values so even a gainclone guy could build it!
Who's gonna make a prototype?
I like the simplicity of your calcs. After all, if we want "everything plus the kitchen sink" we should just get Apox's.
Also, the minimalist approach seems the right way to go with a chip amp. As Peter said, even a few positions can be quite useful.
I'd really appreciate a schematic that is totally clear, very simple, with parts values so even a gainclone guy could build it!
Who's gonna make a prototype?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.