Gainclone Semi-Ultimate and Foreplay III

[bsmith]

I am a novice at DIY. I am just learning electronics. I had previously built a Foreplay III, and want to build a gainclone and use the Foreplay as line stage. It will be ugly (yes, there are some on this site who can only afford a recycled VCR case to enclose a project), but I hope that it will work.

As a newbie to DIY, I don't know the fine points of combining chipamps and tubes. I understand that there are potential problems with this combination. Is there anything that I need to know about?

Thanks for your patience with a newbie.

Bill

 

[bsmith]

Well, my post has been up for a few days and no one has responded. I hope that means that there is no problem with my proposed set up. Some of the information I have been reading has mentioned something about impedance matching problems when using tube pre-amp/line stage. This is a concept that I (very much a newbie) have yet to master. In any case, I will plan on using the Foreplay III, and forgo the hassle of installing two potentiometers.

Thanks to all who read my post!

Bill

 

[Fast Eddie D]

I can't open your gainclone file but I do know what a gainclone is.

The "Foreplay" circuit is capacitor coupled so that avoids one potential problem.

You need to configure your gainclone so it has a fairly high input impedance. 10K is typical for solid state but tube circuits usually employ higher impedances, with good reason. You should configure your gainclone circuit so it's 47K input impedance minimum; 100K would probably be better.

Without viewing your gainclone circuit, I can't make more specific comments. If it's non-inverting, then you can easily configure it for higher input impedance. How high you can go depends on which chip you're using. 47K should not be a problem; but as you go higher the input impedance of the chip will start to affect the circuit.

 

[AndrewT]

office suite opens the .doc file.

You MUST change the circuit.
keep the non-inverting
keep the 100k volume pot, if Foreplay does not have a vol pot. You don't need two.
add a DC blocking capacitor between the wiper and the 680r. Try 1uF
add an RF attenuating capacitor across R1. Try 1nF
Change R1 to 100k
Add a DC blocking capacitor between Signal Ground and R5. Try 220uF
Add proper decoupling to the chip Power Pins. 100uF electro AND 100nF ceramic X7R to BOTH Pins.

You may have to add a disconnecting network into the link joining Signal Ground to Power Ground.

Check the output offset during the build.
You may find that the mismatch between R3=22k and R1=100k gives too high an offset. This is cured by increasing R3 (& R5) up to nearer R1 value. Keep the R3:R5 ratio the same for the same gain. Higher value for R5 allows a lower value for that 220uF cap.

This is one of the few schematics that show the Signal Ground separate from the Power Ground using different symbols. This shows that the Designer is awake.

The four triangle ground symbols on the left are ALL Signal Ground. These ALL get connected together. The traces MUST run along side and be close coupled to the Signal Trace they are serving.

 

[bsmith]

Thanks Fast Eddie. I have just enough understanding of electronics that I could implement your recommendations if I took the time and studied them. It gives me something to try.

Thanks AndrewT for your comments. Your comments went a bit over my head. I was able to grasp the fact that my circuit needs some major modifications. This surprises me, as the schematic is from the e-bay seller of the $19.95 gainclone kit and PCB (without the LM3875's). It was my hope that I could put a transformer, bridge rectifier, a signal source and a few other things in front of the completed kit, and then I would have a working amp. Are the mods needed: A) to improve the sound, or B) fix problem inherent in the kit or C) fix the problems created by having the Foreplay III as line stage or D) all the above? Sorry for this question; I am a newbie just learning both electronics and DIY audio.

Thanks again to both of you for your responses.

Best regards, Bill

 

[AndrewT]

The circuit is closely based on a "stripped down" implementation in the National Datasheet.
This version can ONLY be recommended to experienced builders who are able to debug, if any problems come to light.
Debugging requires expertise and a well resourced (amateur) work shop.

It is very likely that this Ebay supplier provides NO CUSTOMER SUPPORT.
You will be on your own, when it misbehaves or worse Blows up !

The second paragraph is compulsory to eliminate potential operational misbehaviour.
The following paragraphs are optional, I recommend them to you.

 

[Mark Whitney]

I think it is this AMP:

http://www.diyaudio.com/forums/chip-amps/209960-i-designed-gc-lm3875.html#post2968775

 

[AndrewT]

typical of a Far East cloner.

LJM does give some customer support.
But we still see many of his "circuit" builders coming here to ask us to solve their problems.

 

[Mark Whitney]

What about the North American cloners. The builders of these kits also seem to find their way here.

 

[bsmith]

Thanks again for all your comments.

I wanted to keep the design and implementation simple, as this is my first "real" (AKA non-Bottlehead) DIY project. Especially since my electronics "expertise" is virtually non-existent.

Mark, I checked your "I designed the GC LM3875" thread and found that PEMO (comment #36) had built the unit pretty much stock. Is there a problem with his implementation? I was hoping to do the same.

Thanks again for your help for this newbie.

Best regards, Bill

 

[Mark Whitney]

I have not build this kit so its difficult to make a judgement on its quality.

For the first time kit builder it does have the advantage of being an all in design. PSU, AMP and speaker protection.
It removes the added complication of a separate PSU and not having to design a grounding scheme.
For the more experienced builder there is not much room for experimentation or upgrading.

If it is any good with your pre-amp, I could not say. Only one way to find out.

 

[Fast Eddie D]

I think it is this AMP:

http://www.diyaudio.com/forums/chip-amps/209960-i-designed-gc-lm3875.html#post2968775

- No RF input filter
- No capacitor in feedback path and no servo - what's offset going to be like? You have to have one or the other.
- No snubber or protection of any kind on the output
- No input cap

It looks like a lot of things could go wrong to me.

 

[Mark Whitney]

That is why its called a GainClone and not a ChipAmp.

 

[bsmith]

Looks like I will need something besides just a toroidal transformer in front of my completed PCB. I have more research/work to do.

Thanks for the head's up.

Bill

 

[bsmith]

As I have said repeatedly, I am a newbie at DIY. So my comments may be way off base.

I found on the web instructions on how to build a gainclone with a LM3886 chip:

DIY HiFi Gainclone Power Amplifier (2x68w, Class AB-A, LM3886)

The circuit looks somewhat similar to the one I am making – it is my understanding that the ways to build a LM3886 amp and a LM3875 amp are very similar.

According to the above schematic, the items mentioned by Fast Eddie are not needed – with the possible exception of input capacitors before the rectifier (maybe I need then to prevent voltage spikes coming in from the mains?).

This is the critical question. Will the amp that I am building: a toroidal transformer, optional input caps, followed by the completed PCB, work?

Thanks again for your input, and your patience with this newbie!

Regards, Bill

 

[AndrewT]

still too many missing components.

 

[sfthurber]

A few thoughts IMHO related to the vacuum tube aspects of your setup. Even though the 12au7 is a medium mu device, the cathode follower in the second preamp stage will have quite a low output impedance and very good current delivery, by vacuum tube standards, so there is not much to worry about the gainclone loading it excessively with any reasonable input circuit design. There are a couple of things you can do in the preamp to lower its noise floor so as better to exploit the gainclone's low noise floor. 1-change the input pot to a shunt-style setup to get its noise largely out of the signal path; 2-replace both stop resistor with ferrite beads to get their noise out of the signal path; 3-replace the diode bias with a constant current source; 4-check whether the second stage really needs to have its heater elevated to stay within the 12au7 cathode's relatively high breakdown limit -- because from a noise control standpoint, elevating the heater is quite a terrible idea. See, e.g. DIY 12AU7 Tube preamp for a very low noise 12au7-based preamp design that has a non-elevated stage 2 heater even though the cathode voltage is quite high.

 

[Fast Eddie D]

• According to the above schematic, the items mentioned by Fast Eddie are not needed – with the possible exception of input capacitors before the rectifier (maybe I need then to prevent voltage spikes coming in from the mains?)

  
  Do you know enough to determine for yourself whether these items are necessary, or not?
  
  Do you know enough to determine whether a circuit is malfunctioning because you chose to leave these items out, based on a circuit you saw on the web?
  
  Do you know enough to determine the changes in specs, measurements, and performance that may (in most cases probably will) result from the omission of these items? For example, suppose you build the circuit sans the "unnecessary items" and it measures 1 volt output offset. Why did this happen? What steps are you going to take to bring this specification into range?
  
  Do you wonder why circuits designed for consumer and pro use almost always include these "unnecessary items" despite bean counters pressuring the design engineers to save a penny, or even a tenth of a penny, per circuit?
  
  The fact is that these stripped down circuits can be made to work on the bench, and they will often work without a hitch when assembled and used in a system. The addition of these "unnecessary items" makes the operation of the completed device much more foolproof. In other words, you might hook up your newly assembled amplifier to one sound system and it's all good. But then you mix the components up (new preamp, speakers, whatever) and all of a sudden your amplifier isn't performing the same. And since you built the thing, it's going to be up to you to figure out why.

 

[Bare]

Curious re the fixation on the Gainclone/stripper version and Not a Chipamp? Coin issues?
3886 chip amps are easily found, either as bare chips, kits or built and range from very inexpensive to yikes.
Plus the few I'm familiar with, do not need/benefit from a 4play as llinestage/pre
I have all the chip amps I currently need.. just wondering is all.

 

[sfthurber]

I found on the web instructions on how to build a gainclone with a LM3886 chip:
DIY HiFi Gainclone Power Amplifier (2x68w, Class AB-A, LM3886)

if all the missing pieces were not enough to worry you, note also that the instructions call for a 25-0-25 transformer, which will push the PSU caps roughly 40% over their absolute maximum 25 volt rating and will also push the suggested 50 volt rectifiers to roughly 40% over their absolute maximum ratings.