Konnichiwa,
Psooible - sure. But why?
HOW ABOUT THIS:
Unbalanced Input using a split load (aka Concertina) phasesplitter. Uses 1/2 ECC88 (or whatever you care to choose), when driving class A1 output stage balance is determined by the cathode and anode resistor.
Followed by:
Balanced input, using bootstrapped current source loaded cathode follower, using 2 X ECC88 (or whatever you care to choose) per phase as bootstrapped follower and cascoded current source.
Then follow with your bridged/paralell LM3875/3886.
If you don't like Valves, all this can be done with FETS or IC's (as buffers) nearly as well.
I normally despise adding a complication just to do so. Want to do hybrid? Add your Valves so they actually do something....
How about going past Rassmussen San's design and use the Valve as (inverting) gainstage [multiple cascoded Mu Stage - highly linear - I call it "Super Mu Stage]? And how about operateing the output IC(s) as near unity gain inverting "current follower(s)" with a oncertina input splitter where misguided audiophiles insist on using an RCA inputs instead using workable XLR's?
Sayonara
Elkaid said:
Psooible - sure. But why?
HOW ABOUT THIS:
Unbalanced Input using a split load (aka Concertina) phasesplitter. Uses 1/2 ECC88 (or whatever you care to choose), when driving class A1 output stage balance is determined by the cathode and anode resistor.
Followed by:
Balanced input, using bootstrapped current source loaded cathode follower, using 2 X ECC88 (or whatever you care to choose) per phase as bootstrapped follower and cascoded current source.
Then follow with your bridged/paralell LM3875/3886.
If you don't like Valves, all this can be done with FETS or IC's (as buffers) nearly as well.
I normally despise adding a complication just to do so. Want to do hybrid? Add your Valves so they actually do something....
How about going past Rassmussen San's design and use the Valve as (inverting) gainstage [multiple cascoded Mu Stage - highly linear - I call it "Super Mu Stage]? And how about operateing the output IC(s) as near unity gain inverting "current follower(s)" with a oncertina input splitter where misguided audiophiles insist on using an RCA inputs instead using workable XLR's?
Sayonara
Thank you,
The sole purpose of my question was to understand a little bit more the role of each parts in the schematic. Ok, it may seems stupid or senseless ideas for experienced people but everyone starts from somewhere eh ?
I asked the previous questions with one goal in mind :
I wanted to try my hand on the paralleled/bridged configuration using input valves as seen on Joe Rasmussen's schematic but without the input transformer. I wanted to do it on my own but I'm feeling annoying sometimes with my questions. Sorry if it is my bad english that makes me "interpreting" too much
Thanks for your help
Regards,
The sole purpose of my question was to understand a little bit more the role of each parts in the schematic. Ok, it may seems stupid or senseless ideas for experienced people but everyone starts from somewhere eh ?
I asked the previous questions with one goal in mind :
I wanted to try my hand on the paralleled/bridged configuration using input valves as seen on Joe Rasmussen's schematic but without the input transformer. I wanted to do it on my own but I'm feeling annoying sometimes with my questions. Sorry if it is my bad english that makes me "interpreting" too much
Thanks for your help
Regards,
Elkaid said:
Hi Elkaid
Will it work? Will the original circuit I posted work? I haven't actually built it yet but hope I get the go-ahead. I have run simulations that says it should work but there may be gremlins that need to be sorted out when hands on work starts.
Re the Tx, yes, in theory it could be replaced by IC1 as per your amended circuit. I will put my money on the tx as far as sonic pay-off is concerned. The Jensen is expensive but maybe not so expensive based on what it delivers.
Using IC1 also opens the door on what power supply you're going to use on the +/- 15V rails. These will be critical.
Possibly look at using LM431Z shunt regulators? Use a fet to current send the shunt with about 12-15mA. Voila! A version of the VSE SuperReg.
JR
PS: Using 0.1R (shown in the first schematic) instead of 0.22R might be dodgy. Rowland uses 0.22R and the simulated circuits worked fine. It's a safe option as we know that value has been used here before.
Konnichiwa,
Well, as said, if you do not want to use a transformer he best phasesplitter (if buffered) is the concertina. If you don't know what a concertina phasesplitter is you could google it....
One site you might not directly come up, as it uses the term "cathodyne" is this
http://www.bonavolta.ch/hobby/en/audio/split.htm
A circuit suitable in principle would have a normal cathode follower, input directly to grid, 100K gridleak resistor, but with a positive supply of twice the value of the negative supply and a resistor equal in value to that of the cathode resistor in the Anode circuit.
The signal is taken from anode (inverted) and cathode (non-inverted) via coupling capacitors. The signal from the Anode is ideally buffered by another cathode follower, grid to anode of input valve and anode to the positive supply which is twice the negative supply. The cathode resistor then needs to be twice the value of the resistors used in anode/cathode of the input valve.
I would use a mains transformer with 55V + 55V Windings, giving around -75V and +150V when a voltage doubler is used for the positive voltage. If we use the 6922 and run 12.5mA per section we need resistors of 6K2 in anode and cathode of the splitter plus a 12k Cathode resistor for the follower buffering the anode output from the phasesplitter. Make all resistors 5 Watt types.
That should do the trick.
Sayonara
Elkaid said:
Well, as said, if you do not want to use a transformer he best phasesplitter (if buffered) is the concertina. If you don't know what a concertina phasesplitter is you could google it....
One site you might not directly come up, as it uses the term "cathodyne" is this
http://www.bonavolta.ch/hobby/en/audio/split.htm
A circuit suitable in principle would have a normal cathode follower, input directly to grid, 100K gridleak resistor, but with a positive supply of twice the value of the negative supply and a resistor equal in value to that of the cathode resistor in the Anode circuit.
The signal is taken from anode (inverted) and cathode (non-inverted) via coupling capacitors. The signal from the Anode is ideally buffered by another cathode follower, grid to anode of input valve and anode to the positive supply which is twice the negative supply. The cathode resistor then needs to be twice the value of the resistors used in anode/cathode of the input valve.
I would use a mains transformer with 55V + 55V Windings, giving around -75V and +150V when a voltage doubler is used for the positive voltage. If we use the 6922 and run 12.5mA per section we need resistors of 6K2 in anode and cathode of the splitter plus a 12k Cathode resistor for the follower buffering the anode output from the phasesplitter. Make all resistors 5 Watt types.
That should do the trick.
Sayonara
Banned
Joined 2002
Banned
Joined 2002
Help Please!!!!!
I just finished building Joes Hybrid and I'm having all sorts of trouble. Can someone double check my work. Any help is appreciated.
I am using the circut pretty much verbatum. When I power up the amp I get a loud constant hum (noise) that doesn't change with volume and the tube doesn't light up at all.
Pin connections on 6922 tube.
6922 Schematic
1 and 6 are connected togeth getting +35v
2. is connected to the middle tab of the pot through a 3k9 resistor
3 and 8 are connected to the -35v supply
4 and 5 are getting 6.3v (closer to 8v) and are connected together.
7 is the same as 2
9 no connection
There is close to half a volt of dc on one speaker terminal and about 300mv on the other.
I have a minimized gc working but I always run into trouble with tubes.
Thanks for any help...Lee
I just finished building Joes Hybrid and I'm having all sorts of trouble. Can someone double check my work. Any help is appreciated.
I am using the circut pretty much verbatum. When I power up the amp I get a loud constant hum (noise) that doesn't change with volume and the tube doesn't light up at all.
Pin connections on 6922 tube.
6922 Schematic
1 and 6 are connected togeth getting +35v
2. is connected to the middle tab of the pot through a 3k9 resistor
3 and 8 are connected to the -35v supply
4 and 5 are getting 6.3v (closer to 8v) and are connected together.
7 is the same as 2
9 no connection
There is close to half a volt of dc on one speaker terminal and about 300mv on the other.
I have a minimized gc working but I always run into trouble with tubes.
Thanks for any help...Lee
Re: Help Please!!!!!
The tube not lighting up should be your first error to track down. Are you certain that you have 6.3V across pins 4 & 5. If you 8V this could damage the tube.
Check with your meter that you don't have an open circuit in the filaments across same pins.
A couple of things to check:
1) Pins 3 & 8 should be connected via 10K resistors. Have you? You are not clear about that and above makes it sound like you have connected it straight to -35V?
2) One simple error that can be made by first time tube users: Reversing tube connections so that pin 1 has become pin 9. That could also explain why filaments are dead. Look at the bottom of the tube direct, then pin 1 is the first pin (from the gap) and the others are counted clockwise. Check it. You may have filaments across pins 5 & 6 instead of pins 4 & 5.
Simple mistake, but I have seen this happening before.
Disconnect the GC stage and sort out the tube stage first. When it works correctly then the filaments will light up, and you will have about 1V on pins 3 & 8. Also 0V on pins 2 & 7.
Once that is working properly, then wire up the GC output stage.
Let us know how you get on.
Joe R.
king30 said:
The tube not lighting up should be your first error to track down. Are you certain that you have 6.3V across pins 4 & 5. If you 8V this could damage the tube.
Check with your meter that you don't have an open circuit in the filaments across same pins.
A couple of things to check:
1) Pins 3 & 8 should be connected via 10K resistors. Have you? You are not clear about that and above makes it sound like you have connected it straight to -35V?
2) One simple error that can be made by first time tube users: Reversing tube connections so that pin 1 has become pin 9. That could also explain why filaments are dead. Look at the bottom of the tube direct, then pin 1 is the first pin (from the gap) and the others are counted clockwise. Check it. You may have filaments across pins 5 & 6 instead of pins 4 & 5.
Simple mistake, but I have seen this happening before.
Disconnect the GC stage and sort out the tube stage first. When it works correctly then the filaments will light up, and you will have about 1V on pins 3 & 8. Also 0V on pins 2 & 7.
Once that is working properly, then wire up the GC output stage.
Let us know how you get on.
Joe R.
Joe thanks,
I've got the pins in the right order and the -35v are going through the 10k resistors. The only thing off is the 8v on the heaters. On the heater PS what is the best why to drop this. I tried increasing the 5.6 Ohm resistor to 150k, but the drop was nominal.
Thanks for any ideas...Lee
I've got the pins in the right order and the -35v are going through the 10k resistors. The only thing off is the 8v on the heaters. On the heater PS what is the best why to drop this. I tried increasing the 5.6 Ohm resistor to 150k, but the drop was nominal.
Thanks for any ideas...Lee
Hi I have been following this thread for a while and would appreciate someones comments on several questions I have.
Firstly I have tried a few chips now, opa549, opa547 lm3875 and LM3886. I used the same classy Hitano caps from dicksmith
on all clones and cheap metal films and same transformer bridge etc.
The LM3875 sounded best by long shot I thought. Why then are we and Jeff Rowland using LM3886, is it purely power requirement? I know we are free to choose what we prefer, but wondering if this is the preferred chip for bridging?
I also noted that increasing the caps from 1000uF to 2200 really killed the sound, I was surprised how different it sounded. Here in Joes design we are using 10000uF. Would the 1000uF on each chips rails not supply enough current to a bridged amp?
Would Joes circuit with 6 chips per channel work just as well with 4 per channel (understand less power of course) ?
Can the bridged version with tube buffer match the quality of a single chip gainclone like kuei yeui yang wang and Peter have kindly given us?
Firstly I have tried a few chips now, opa549, opa547 lm3875 and LM3886. I used the same classy Hitano caps from dicksmith
on all clones and cheap metal films and same transformer bridge etc.The LM3875 sounded best by long shot I thought. Why then are we and Jeff Rowland using LM3886, is it purely power requirement? I know we are free to choose what we prefer, but wondering if this is the preferred chip for bridging?
I also noted that increasing the caps from 1000uF to 2200 really killed the sound, I was surprised how different it sounded. Here in Joes design we are using 10000uF. Would the 1000uF on each chips rails not supply enough current to a bridged amp?
Would Joes circuit with 6 chips per channel work just as well with 4 per channel (understand less power of course) ?
Can the bridged version with tube buffer match the quality of a single chip gainclone like kuei yeui yang wang and Peter have kindly given us?
Your observations are in line with mine. Some designers probably don't analyze the sound of a chip as much and go for a higher current version. I will probably try bridged/parallel configuration, but with LM3875 chips. Also, it is said that Jeff Rowland amp doesn't sound that well.
>>Can the bridged version with tube buffer match the quality of a single chip gainclone like kuei yeui yang wang and Peter have kindly given us?<<
I was reading comments in other thread and somebody mentioned trying both single chip and parallel setup and single chip sounded better.
I don't see a problem with using just 4 chips per channel.
>>Can the bridged version with tube buffer match the quality of a single chip gainclone like kuei yeui yang wang and Peter have kindly given us?<<
I was reading comments in other thread and somebody mentioned trying both single chip and parallel setup and single chip sounded better.
I don't see a problem with using just 4 chips per channel.
Ok,
One problem down and ? to go. I ended up switching the transformer for the tube and running 6.3 volts AC directly to the heaters. Tube is now lighting up great. But when I plug in the speakers I still have a very loud (speaker damaging )hum that doesnt change at different volume levels.
I have two caps on each chip, when I remove these from the star ground the hum goes away, but can't hear music. I'm using two transfo's for the chips and their grounds are daisy chained to one point where are the amp grounds meet.
Again any help is appreciated...Lee
One problem down and ? to go. I ended up switching the transformer for the tube and running 6.3 volts AC directly to the heaters. Tube is now lighting up great. But when I plug in the speakers I still have a very loud (speaker damaging )hum that doesnt change at different volume levels.
I have two caps on each chip, when I remove these from the star ground the hum goes away, but can't hear music. I'm using two transfo's for the chips and their grounds are daisy chained to one point where are the amp grounds meet.
Again any help is appreciated...Lee
Hi Peter,
Im quite interested in the low power version, single chip with Joes tube buffer. BTW forgot to thank You Joe for your contribution and look forward to building your JLTi
Im also interested in your feedback regarding the Bridged tube buffered amp when you build it, will be looking for your post
Dont have any tubes for now so Im currently working on the bridged version using SSM2142 chip at the front.
Have really enjoyed this chip amp stuff and find them very forgiving, even done some arc welding with my 1000uF caps and chip pins And they dont blow Its also a good and rasonably inexpensive way to try out high end components. I must say I was really sceptical about black gates and high end resistors but they work.
Im quite interested in the low power version, single chip with Joes tube buffer. BTW forgot to thank You Joe for your contribution and look forward to building your JLTi
Im also interested in your feedback regarding the Bridged tube buffered amp when you build it, will be looking for your post
Dont have any tubes for now so Im currently working on the bridged version using SSM2142 chip at the front.
Have really enjoyed this chip amp stuff and find them very forgiving, even done some arc welding with my 1000uF caps and chip pins And they dont blow
Its also a good and rasonably inexpensive way to try out high end components. I must say I was really sceptical about black gates and high end resistors but they work.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.