The gain will increase if you bypass the input triode cathode resistors R104 and R204 with bypass capacitors.
Try 50 uF, 10V. Connect them across the 1k Ohm cathode resistors R104 and R204 (- end to ground).
I am using the schematic at post # 316. Is that the correct one for the parts and connections you have?
Try 50 uF, 10V. Connect them across the 1k Ohm cathode resistors R104 and R204 (- end to ground).
I am using the schematic at post # 316. Is that the correct one for the parts and connections you have?
The schematic at post # 316 should be OK.
Very weak output ...
There have been many mistakes/errors with the RCA selector switch;
Also wiring the volume potmeter is a pitfall.
If you look at earlier posts, builders tend to put a "grid-stopper" resistor in series with the input, 3k3.
Apparantly that surpresses HF oscillitions. If you consider to do that, now is the time:
hook up your CD directly to the new resistor and check volume, should be very loud.
Lastly: if you have volume 100% up, do you hear ANY hum/hiss in your speakers?
Turn of the music first
greets, Marcel
Very weak output ...
There have been many mistakes/errors with the RCA selector switch;
Also wiring the volume potmeter is a pitfall.
If you look at earlier posts, builders tend to put a "grid-stopper" resistor in series with the input, 3k3.
Apparantly that surpresses HF oscillitions. If you consider to do that, now is the time:
hook up your CD directly to the new resistor and check volume, should be very loud.
Lastly: if you have volume 100% up, do you hear ANY hum/hiss in your speakers?
Turn of the music first
greets, Marcel
Hi dmnstewart,
Can you turn up the volume until it's distorted? This will only generate a couple watts (11 best case - which isn't too bad I guess), so it will tend to be low in volume. You can also try a 6SL7 instead of the 6SN7 for higher gain. The EL34 will deliver 40 watts in push-pull configuration.
Too bad they used an EL34 for the output. You would probably be further ahead if the unit was designed for an EL84/6BQ5A (with slightly lower B+) or a 6V6GT. The 6V6GT might just plug right in. Same for a 6L6GC.
Remember, a single ended output stage will always deliver less than 0.5 the power that a push-pull stage will. So a P-P amplifier using 6BQ5 will output more power than this easily. (5.7 watts single ended or 17 watts P-P). 6V6GT will deliver 5.7 watts, or 14 watts P-P. These figures are right out of the RCA tube manual RC-27.
Interestingly, your schematic shows a beam power tube, not a pentode. A 6CA7 is a beam power tube roughly equivalent to the EL34. The EL34 is a tall skinny tube whereas the 6CA7 is slightly shorter, but much fatter. You can find these tubes marked either way, but the aspect ratio of the glass is a dead giveaway as to which it is. I greatly prefer the 6CA7 in P-P applications to the EL34.
I hope that helps, Chris
Can you turn up the volume until it's distorted? This will only generate a couple watts (11 best case - which isn't too bad I guess), so it will tend to be low in volume. You can also try a 6SL7 instead of the 6SN7 for higher gain. The EL34 will deliver 40 watts in push-pull configuration.
Too bad they used an EL34 for the output. You would probably be further ahead if the unit was designed for an EL84/6BQ5A (with slightly lower B+) or a 6V6GT. The 6V6GT might just plug right in. Same for a 6L6GC.
Remember, a single ended output stage will always deliver less than 0.5 the power that a push-pull stage will. So a P-P amplifier using 6BQ5 will output more power than this easily. (5.7 watts single ended or 17 watts P-P). 6V6GT will deliver 5.7 watts, or 14 watts P-P. These figures are right out of the RCA tube manual RC-27.
Interestingly, your schematic shows a beam power tube, not a pentode. A 6CA7 is a beam power tube roughly equivalent to the EL34. The EL34 is a tall skinny tube whereas the 6CA7 is slightly shorter, but much fatter. You can find these tubes marked either way, but the aspect ratio of the glass is a dead giveaway as to which it is. I greatly prefer the 6CA7 in P-P applications to the EL34.
I hope that helps, Chris
Technically, the EL34 started in life as a True Pentode Tube. Technically, the 6CA7 started in life as a True Pentode Tube.
Technically, the EL84 started in life as a True Pentode. Technically, the 6BQ5 and the 6V6 started in life as Beam Power tubes. But manufacturers do whatever they want to do today. So, you can find both pentode and beam power versions of both the EL84 and the 6BQ5. I haven't seen any modern (current production) 6V6s, so I do not know if they still all have beam forming plates (G#3) or not.
True Pentodes have a Suppressor Grid (Grid #3). These consist of 2 vertical grid posts and the suppressor grid wire is wrapped around those posts. Most of these have Grid #3 connected internally in the tube to the Cathode. The EL34 and 6CA7 are noted exceptions.
Most True Beam Power tubes had a pair of Beam Forming Elements, Grid #3 (beam forming plates in parallel). Most of these have Grid #3 connected internally in the tube to the Cathode.
A lot of Beam Power tubes were RF tubes, the 807 is a well known one.
Some tubes are referred to as Beam Power tubes because they have the Control Grid (1)and the Screen Grid (#2) at exactly the same pitch and starting point. That makes for beam forming too. They may or may not have the Beam Forming Plates, or the Suppressor Grid wires for Grid #3.
If you truly have the amp wired as per the schematic post #316, and if the parts are good, you should have plenty of gain with the ECC83.
Gain and power are two different things. One amp can have a voltage gain of 7, and have 12 Watts, and another amp can have a voltage gain of 20 and only have 2 Watts out, etc.
That raises the question, what input tube are you using? The 6SN7 is low gain, the ECC83 is high gain, and the 6SL7 is medium gain. But that all depends on the Cathode resistor, and the Plate load resistor you are using. Each of those tubes requires cathode and plate resistors to be different values for each type of tube. If not, the gain statement I made will not be true, and more importantly the dynamic range and the distortion will not be optimum.
And if you use both triodes wired in parallel for one channel, or only use one triode per channel, you also have to change the cathode and plate resistors.
Changing tubes like that without regard to the cathode and plate resistor values is not a good idea, the results are not the best.
If you want to tube roll, use tubes with similar characteristics, and the correct resistor values. Of course pay attention to the filament requirements, the 12.6 V tubes mentioned below can be wired for 6.3V.
12AX7 and ECC83; 6SN7 and 12AU7; 6SL7. The 5751 is between the 6SL7 and 12AX7/ECC83.
For tube rolling of tubes with similar characteristics, but that are both 9 pin and 8 pin, you will have to wire 9 and 8 pin sockets in parallel, or build a socket adapter.
Be sure that the output tube screen is wired to the screen tap, and that the output tube plate is wired to the plate tap, otherwise you will get low gain, and low power.
Technically, the EL84 started in life as a True Pentode. Technically, the 6BQ5 and the 6V6 started in life as Beam Power tubes. But manufacturers do whatever they want to do today. So, you can find both pentode and beam power versions of both the EL84 and the 6BQ5. I haven't seen any modern (current production) 6V6s, so I do not know if they still all have beam forming plates (G#3) or not.
True Pentodes have a Suppressor Grid (Grid #3). These consist of 2 vertical grid posts and the suppressor grid wire is wrapped around those posts. Most of these have Grid #3 connected internally in the tube to the Cathode. The EL34 and 6CA7 are noted exceptions.
Most True Beam Power tubes had a pair of Beam Forming Elements, Grid #3 (beam forming plates in parallel). Most of these have Grid #3 connected internally in the tube to the Cathode.
A lot of Beam Power tubes were RF tubes, the 807 is a well known one.
Some tubes are referred to as Beam Power tubes because they have the Control Grid (1)and the Screen Grid (#2) at exactly the same pitch and starting point. That makes for beam forming too. They may or may not have the Beam Forming Plates, or the Suppressor Grid wires for Grid #3.
If you truly have the amp wired as per the schematic post #316, and if the parts are good, you should have plenty of gain with the ECC83.
Gain and power are two different things. One amp can have a voltage gain of 7, and have 12 Watts, and another amp can have a voltage gain of 20 and only have 2 Watts out, etc.
That raises the question, what input tube are you using? The 6SN7 is low gain, the ECC83 is high gain, and the 6SL7 is medium gain. But that all depends on the Cathode resistor, and the Plate load resistor you are using. Each of those tubes requires cathode and plate resistors to be different values for each type of tube. If not, the gain statement I made will not be true, and more importantly the dynamic range and the distortion will not be optimum.
And if you use both triodes wired in parallel for one channel, or only use one triode per channel, you also have to change the cathode and plate resistors.
Changing tubes like that without regard to the cathode and plate resistor values is not a good idea, the results are not the best.
If you want to tube roll, use tubes with similar characteristics, and the correct resistor values. Of course pay attention to the filament requirements, the 12.6 V tubes mentioned below can be wired for 6.3V.
12AX7 and ECC83; 6SN7 and 12AU7; 6SL7. The 5751 is between the 6SL7 and 12AX7/ECC83.
For tube rolling of tubes with similar characteristics, but that are both 9 pin and 8 pin, you will have to wire 9 and 8 pin sockets in parallel, or build a socket adapter.
Be sure that the output tube screen is wired to the screen tap, and that the output tube plate is wired to the plate tap, otherwise you will get low gain, and low power.
Last edited:
@6A3sUMMER Thanks a bunch for the penthode - tetrode - beams explanation. I was totally puzzled to see EL84 and also EL34 modelled in LTSpice as tetrodes ... it now makes totally sense. Also I remember that the beam cathode was an US invention driven to circumvent Dutch Philps patents for the Pethode design. It may be I remember it wrong because Iam Dutch
If Steward sticked to the schematic the EL34 amp must be loud. Heavy output transformers, hot tubes ... My A9 is fully "stock" plus supressor resistor and is VERY loud. Far too loud so I consider switching to triode mode.
If you want to swap tubes, why not test-swap with a simulations Help offered here. I'll post a schematic / simulation of the 2N9PJ-EL34 SE this week. Current version hits 15 watts power @8ohms but I need to refine output transformer. If you want it sooner PM me, I have not checked anything on errors, just have been trying to get the darn thing working, it does now as wished for But may have BIG errors so on request only. Anyone experienced in ltspice interested in checking my LTSpice study material?
Happy listenings, Marcel
The simulation can be used to test / evalute modifications.
If Steward sticked to the schematic the EL34 amp must be loud. Heavy output transformers, hot tubes ... My A9 is fully "stock" plus supressor resistor and is VERY loud. Far too loud so I consider switching to triode mode.
If you want to swap tubes, why not test-swap with a simulations
Help offered here. I'll post a schematic / simulation of the 2N9PJ-EL34 SE this week. Current version hits 15 watts power @8ohms but I need to refine output transformer. If you want it sooner PM me, I have not checked anything on errors, just have been trying to get the darn thing working, it does now as wished for But may have BIG errors so on request only. Anyone experienced in ltspice interested in checking my LTSpice study material? Happy listenings, Marcel
The simulation can be used to test / evalute modifications.
A9
I went through the whole circuit with my Fluke multimeter and found out I had incorrectly used the 100 ohm resistor on the drive tube plate/anode and the 1kohm between the linear tap and the el34 screen. Should have double checked! All working very well now. No hum or hiss. Overall very pleased with the build standard apart from the wide tolerance of the tube bases. Will now try out some of the modifications on this thread.
Thanks everyone
I went through the whole circuit with my Fluke multimeter and found out I had incorrectly used the 100 ohm resistor on the drive tube plate/anode and the 1kohm between the linear tap and the el34 screen. Should have double checked! All working very well now. No hum or hiss. Overall very pleased with the build standard apart from the wide tolerance of the tube bases. Will now try out some of the modifications on this thread.
Thanks everyone
Great moment for Stewart, having fun with more db's
Tube bases are bad indeed, Good thing the A9 is not a guitar amp, tubes would be rolling around.
HINT FOR NEW BUILDERS: exchange the tube holders BEFORE soldering And also sometimes the sockets/pinholders are just not twisted enough to sit properly and tight in the porcelain. That extra twisting is also easy to do before soldering ... have fun.
Tube bases are bad indeed, Good thing the A9 is not a guitar amp, tubes would be rolling around.
HINT FOR NEW BUILDERS: exchange the tube holders BEFORE soldering
And also sometimes the sockets/pinholders are just not twisted enough to sit properly and tight in the porcelain. That extra twisting is also easy to do before soldering ... have fun.A9
Works very well on my NAD cd player. Very little hum. Have to use a preamp(NAD) for my Technics record player (magnetic cartridge)) and my REGA 78 rpm turntable (magnetic pick up) But results are very good despite some mis-match. Have put in bypass capacitors on input valve/tube, with increase in output with some very low hum. Nothing that affects listening quality. Will now enjoy music before any more modifications. Current speakers which are Mission/Denon 88 Db soon to be replaced by B & W speakers.
Works very well on my NAD cd player. Very little hum. Have to use a preamp(NAD) for my Technics record player (magnetic cartridge)) and my REGA 78 rpm turntable (magnetic pick up) But results are very good despite some mis-match. Have put in bypass capacitors on input valve/tube, with increase in output with some very low hum. Nothing that affects listening quality. Will now enjoy music before any more modifications. Current speakers which are Mission/Denon 88 Db soon to be replaced by B & W speakers.
power output low
Some what puzzled .Ive just built one of these , it works - but with only a little power, 2 - 3 watts. Have fitted the by pass 470uf , grid resistors , and 100r ground resisters on the 6.3v heater . Currently using the supplied tubes. Both channels are the same . All seems good , except that i only get a max of 6v peak on the 8r output loaded with a 8r2 10w resistor , above this the positive clips and the negative goes to about 8v before clip. If i move the resistor to the 4r output i can get 9v peak before clip , the 8r terminal shows 11v on scope. If increase the load to 4r on the 4r terminal i get a max of 6v before clip. The signal on the 6n9pj is a clean sine wave,even when the output is virtually a square wave. ( i use 600 - 800 hz for testing.) Ive checked and rechecked the connections and measured all the values , all looks good. Transformer wires go to the correct places. Ive measured the voltages and compared to those on page 32 and there are differences , i seems to me that there is too little voltage on the output transformer. At the 39r resistor 342v , B1 - 330v , EL34 pin3 324v , pin4 329v , pin1/8 23v. right channel is similar. that puts 6vdc across the output transformer , compared to the measurements on page 32 that one has 10v . As the dc is a little higher i would have expected arround 11v. Does any one have any ideas whats wrong here? Also, Was expecting the transformers to get a bit warm , but they dont.
Some what puzzled .Ive just built one of these , it works - but with only a little power, 2 - 3 watts. Have fitted the by pass 470uf , grid resistors , and 100r ground resisters on the 6.3v heater . Currently using the supplied tubes. Both channels are the same . All seems good , except that i only get a max of 6v peak on the 8r output loaded with a 8r2 10w resistor , above this the positive clips and the negative goes to about 8v before clip. If i move the resistor to the 4r output i can get 9v peak before clip , the 8r terminal shows 11v on scope. If increase the load to 4r on the 4r terminal i get a max of 6v before clip. The signal on the 6n9pj is a clean sine wave,even when the output is virtually a square wave. ( i use 600 - 800 hz for testing.) Ive checked and rechecked the connections and measured all the values , all looks good. Transformer wires go to the correct places. Ive measured the voltages and compared to those on page 32 and there are differences , i seems to me that there is too little voltage on the output transformer. At the 39r resistor 342v , B1 - 330v , EL34 pin3 324v , pin4 329v , pin1/8 23v. right channel is similar. that puts 6vdc across the output transformer , compared to the measurements on page 32 that one has 10v . As the dc is a little higher i would have expected arround 11v. Does any one have any ideas whats wrong here? Also, Was expecting the transformers to get a bit warm , but they dont.
Hi easygoing
It seems to me, that the screen grid is to high (329v on EL34 pin 4). Check the wirering according to what 6A3Summer wrote
The A9 has normally a screen tap at about 50%, so the voltage shut be about 325v.
If the wirering is correct, check the output transformers according to what ygg-it wrote in his post #28.
It seems to me, that the screen grid is to high (329v on EL34 pin 4). Check the wirering according to what 6A3Summer wrote
The A9 has normally a screen tap at about 50%, so the voltage shut be about 325v.
If the wirering is correct, check the output transformers according to what ygg-it wrote in his post #28.
Thanks for reply, had more time today, let it warm up properly. Secondary resistance seems good 0.9R on 8R and 0.6R on 4R. On primary i have - red to blue 99R , blue to brown 114R , red to brown 212R. with 100Vac @ 600hz across primary i get 3.41v on 4R and 4.52V on 8R outputs. which seems about right. Measurement and connections on EL34 - 1+8 25.2v 2x1k and 220uF all to ground, 2+7 6.2Vac , 3 325v brown wire , 4 330v 100R to blue wire, 5 32mV 330k to ground and coupling cap, 6 not used. Tried swapping red and brown wires , made no difference. Output still the same as it was, ie 6v at cippling on 8R.
The El34 cathode (pin 8) is at 25V. The current over the 500 ohm cathode resistor looks to be =25/500R or 50mA. This match the EL34 "Triode" mode curve and not the expected UL mode. Your output power also seems to match that of a SE EL34 in triode mode. I wonder if your EL34 tube screen is some how blown open? Have you try switch tubes (ie. between left and right channels)?
Put a scope probe on both pin 4 (screen) and pin 3 (Anode) and compare the AC signals. What percents of pin 4 signal is compared to pin 3? Are the the signal in phase? EL84 transformer are higher impedance. Your measurements (plate and speaker) indicates around 4K. Not untypical for EL34 use.
Last edited:
That may be "working condition" for this amp. I see a manufacturer's rating of 12W at 1% THD. This amp is ultra-linear with no feedback. These are mutually exclusive conditions, you are being lied to! I would wonder if you could get more than 1W at 1% THD. This is very standard practice for Chinese amps; no surprises here.
The driver stage has a ton of headroom.
Those are dead nuts on in terms of what the schematic shows, other than slightly high screen voltage.
What's important is the DC current across the cathode resistor, which you have measured to be correct.
The power transformer looks to be adequately sized for what it's doing. It should warm up after a few hours of use. As you've said before, you have about 10V across the OPTs at about 45mA, which is less than 1W of power dissipated. They aren't going to get warm.
You bought a $200 tube amp from China, YGWYPF. Try triode strapping it.