• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Olson stereo amps & docs ?

david winter

david winter

Classical music lover.
Hi,

It looks like there's little technical information about Olson tube amplifiers.
I'm trying to go through an unidentified model equiped with a 5AR4 rectifyer, four EL84 PP and 5 12AX7 (all testing good).
There's some residual audible ripple after recaping the power supply whatever the position of the volume control. A NOS 5AR4 didn't help, neither another set of EL84.
Components are essentialy Suzuki branded and the wiring is typical of a kit.
The two 12AX7 from the phono / tape head entries are DC powered from the bias circuit of the EL84.
The amplifier I'm working on doesn't have exactly the same housing / front place as this specimen but the chassis looks identical.

Any infos ? Thanks !
 
Register to hide this ad
david winter said:
Hi,

It looks like there's little technical information about Olson tube amplifiers.
I'm trying to go through an unidentified model equiped with a 5AR4 rectifyer, four EL84 PP and 5 12AX7 (all testing good).
There's some residual audible ripple after recaping the power supply whatever the position of the volume control. A NOS 5AR4 didn't help, neither another set of EL84.
Components are essentialy Suzuki branded and the wiring is typical of a kit.
The two 12AX7 from the phono / tape head entries are DC powered from the bias circuit of the EL84.
The amplifier I'm working on doesn't have exactly the same housing / front place as this specimen but the chassis looks identical.

Any infos ? Thanks !
Click to expand...
All the Olson amps I've seen are Japanese and their components don't seem to hold up well - especially the gray capacitors. You should plan on replacing every one of them.

It would help if you posed pics of the unit you're actually working on.
 
Agree, get those Suzuki PIO caps out of there. They could be leaking DC from one stage to the next. Might even be part of your hum issue.
 
I have done a number of tube amps including some Japanese ones (Trio, Sansui and the like).
I changed "those" six 0.02 plate capacitors between 12AX7 sections as well as "those" four 0.05 between the 12AX7 drivers and EL84 outputs.
Same with all cathode electrolytics.
All others are located in the Eq circuit and there is no DC there, so I'd be quite surprised that they be the cause of the trouble.
I also adjusted the hum pot. Sounds better in the middle position.
Still that residual 100Hz ripple. Of course I could add a big fat 100µF in parallel to the first 50µF but that's the dirty solution which only reduces the problem instead of fixing it.
 
The underside picture in the link you provided is blurry. Any chance you could take a clear photo of the insides of your unit and post it here? Someone might see something for you.
 
Another thing to know about push-pull amps is they rely on equal current draw through the output transformers to cancel a lot of hum. The B+ tap feeding the output tubes tends to be pretty noisy, so if you have seriously mismatched tubes, or leaky caps causing significantly uneven current flow, it will tend to give you 100 hzhum through the speakres.

ground loop or heater line problems will be mains frequency, so 50hz. Would be good to know which you have.
 
david winter said:
Of course I could add a big fat 100µF in parallel to the first 50µF but that's the dirty solution which only reduces the problem instead of fixing it.
Click to expand...
The data sheet for the 5AR4 rectifier specifies a maximum first cap value of 60uf, so you definitely don't want to add 100uf to the existing 50uf cap.

So did you replace the original PS can capacitors or is that 50uf the original? If original and you have hum, that would be the first place to look.
 
The first thing I did was replacing the electrolytics.
Originals were 40+40 under 350/385V and dry. I used 50+50 under the same voltage.
 
Last edited:
david winter said:
The first thing I did was replacing the electrolytics.
Originals were 40+40 under 350/385V and dry. I used 50+50 under the same voltage.
Click to expand...
Noise issues can be caused by so many different things so you need to work through different areas. I noticed that there are at least 4 places where grounds were made with the chassis. This can cause a ground loop and you might want to redo the grounds if you can not get rid of the hum. Poor wiring layout can place an AC wire, like the filament wires, too close to a DCV wire which may induce hum into the signal circuit
IME with the vintage japanese tube amps of this era you need to be aware of some "weak spots" that can cause noise. One can be the resistors. These are wire wound around a ceramic core which is excellent but the end caps are a pressfit or maybe there is a weak spot weld. But when rebuilding if you manipulated them to change other components that can loosen the end caps and make a poor or noisy connection. The other area that you need to be careful with is all of the solder lugs including the tube socket solder lugs. The metal for the solder lug is slightly thinner than the usual and will break easily if you tug on it when trying to remove wiring or a capacitor lead. The lug can be broken but still attached which can lead to noise issues. Anyway just some things to keep in mind . Good luck.
 
Thanks for your answers. I already removed some black ground wires visible in the photo.
One wire was passing through the whole chassis to reach two power resistor grounding either the headphone or speaker terminals (whichever isn't used). As there was a ground point near those resistors, the wire was considerably shortened.
The black wires in the EL84 cathode resistor circuit were removed as well. The 1K cathode resistor was already grounded (as the mains power transformer is), so the black wire crossing the chassis was useless.
Even with these two mods the problem persists in both channels. I'll investigate more tomorrow.
I just hope the problem is not due to a mains transformer originally designed for 60Hz which radiates a lot more under 50Hz.
 
david winter said:
The black wires in the EL84 cathode resistor circuit were removed as well. The 1K cathode resistor was already grounded (as the mains power transformer is), so the black wire crossing the chassis was useless.
Click to expand...
I checked your photo closely and it appears to me that the cathode resistors are the 2 black ones that are on either side of the big green 1K resistor, which i believe is the voltage dropping resistor for the suppressor grid on the el84. And it looks to me that the yellow wires on the el84 sockets are the ones that go to the cathodes on the el84 sockets which is pin 3. And looking at the top of photo where the black cathode bias resistor is supposed to connect to the solder lug, it appears the the resistor lead is unconnected. So make sure that the 2 black bias resistors are grounded preferably to the same ground as the high voltage or B+ ground.
 
This is a picture effect. Both black 150 ohms cathode resistors are well connected. The remaining solder joints on their pins (to be removed later) correspond to the points where the 47µF cathode electrolytics were originally connected.
The cathode circuit of the EL84 is used to DC-power the two phono/tape 12AX7. In each channel, both EL84 cathodes go to the 150ohm resistor trough a yellow wire. The other ends of both 150 ohm resistors go to what I call the "blue point" as per blue wires, from which the filaments of two 12AX7 are powered. At this point, there is the big 1K resistor with a 100µF parallel smoothing capacitor (Nichicon PW radial in this case).
Note that each 470K grid resistor of the EL 84 go to the "blue point" as well. Not sure what this does as 1K is negligible.
 
probably tickles some positive voltage back to the grid so the grid-cathode voltage is where it needs to be. Fair bet your cathode will be somewhere close to 22 volts, and the grid of the output tubes maybe around 10 volts. Grid-cathode voltage ends up around 12, which is ballpark typical for a 6bq5 push-pull output stage at reasonable voltages.

the 470k are your grid to "ground" resistors, but in this case they tie to a voltage source rather than ground for biasing reasons.
 
I found that one output transformer had an open primary on one side and was very resistive one on the other.
I replaced it with an Audax TU 101. Residual hum didn't disappear for such.

Cathodes are around 28-29 volts, grids around 17 volts. Grid-cathode is around 10-11V.
Which gives the 18V for both 12AX7, which makes sense since 24 x Sqr(2) is almost 17.

Still wondering where the problem is...
 
not sure what square roots have to do with it. If the cathodes run to ground through the 12ax7 heaters, thats 14 volts on each tube rated for 12.6, unless there is another resistor somewhere in the chain to drop it down so each tube only gets ~12v

either that or its just biased way wrong, which is also not impossible. Unfortunately without a schematic I'm sort of guessing about whats going on based on description and assumptions that may not be accurate.
 
At what I call the "blue point" there's 18V. This feeds the two 12AX7 heaters which are all connected in series. 24Vac would be required to heat them both, but here a DC voltage is used, in which case the RMS value should be considered, hence the approximately 17V DC. Just as an experience, test one 12AX7 using 12Vac as usual. Then test it with 12Vdc at the filament and compare. You should get the same result as if you heated it with approximately 17Vac (not to be confused with the 17V dc in the above comments).
 
They are rated for 12.6 volts, DC or AC RMS doesn't matter. The heaters don't have different ratings depending on the type of power supplied. If you feed a 12ax7 17 vac it'll emit like crazy until it fries the heater.

Not uncommon for heaters to be starved some for noise reasons, but under 9 volts each seems excessive.
 
You must log in or register to reply here.
Back
Top Bottom