| Aikido warm-up hum, resistor types, etc. - Click HERE for Original Thread |
| cbutterworth |
I just can't stay away from this forum, can I?
The music that I hear from my Aikido / AKSA combo is extremely pleasing. However, I am still having a few hum issues. These can be characterized as:
1. Once the B+ has been switched-on (following 30 seconds of heating), I usually switch on my main amp - AKSA55N a solid-state class AB push-pull. When I do this, I get a very audible 60Hz hum (at least it corresponds better to a 60Hz PC generated tone than a 120Hz tone). After a few minutes, it seems to die away to barely audible. It could be that a greater ground lift resistance (currently 20 ohms) could reduce this further. It does not seem to be the AKSA because if I leave the Aikido switch-on for a few minutes, then switch on the AKSA, I do not get the louder hum. It seems that something maybe warming-up in the preamp.
2. After playing some music - maybe a CD or so, the hum seems more noticeable. This could simply be psychological, but it makes me feel pathological :bawling: only joking.
I have resoldered all pads on the Aikido PCB, except for C5 which is mounted flush with the PCB, and I suppose that I should unbolt the PCB and resolder it for peace-of-mind.
The PSU filter capacitors are large 45uF motor-run caps bought from a reclamation dealer. They appear to be in decent condition (no dings, etc.). Could these be contributing to the noise, and as they warm (if they actually do) the hum level is reduced? I have 470K bleeders across these caps so that I can poke around without risking shock from them. If fully discharged, might they take a short while to fully charge and reach full capacity?
I used Vishay-Dale metal film resistors throughout. Could these generate some noise? If so, what sort of frequency would this be? I did read somewhere that carbon-comps are better for the grid stoppers.
My B+ delay uses an Amperite 30 second tube, which then switches a coil relay that simultaneously switches on B+ while removing power from the Amperite. Could the coil relay (Omron) be causing noise? If so, what kind of frequency?
My signal ground is only using a single run of cat5e. Should I use more substantial wire? I expect that I should as V = IR, so a heavier gauge cable should result in lower V?????
I feel very relieved at least that I am in the tweaking stage and can actually thoroughly enjoy listening.
Thanks,
Charlie |
|
|
| kmtang |
What are the tubes you used in your Aikido amp?
Is the filament voltages are lifted to approx +80Vdc?
Johnny |
|
|
| cbutterworth |
Johnny,
I am using 6SN7's throughout. To be precise, I have Sylvania 6SN7GTA and Tung-Sol 6SN7GTB, although I have also used some Magnavox tubes.
It may be that some of the hum is comming from the Sylvania tubes. I only got the Tung-sols today from an Ebay sale. I put them in the input with the Magnavox in the output. The hum-level was relatively low, even on start-up. When I put the Sylvania in the output, the hum was louder on start-up.
The heater is AC and is referenced to 1/4 B+ which is about 68V based upon 270 B+.
My Sylvania's test well - 2600 for each section (or thereabouts). The Tung-sols about 2100 for each section.
Regards,
Charlie |
|
|
| tube-lover |
hi cbutterworth,Johnny & kmtang,
why not considerate 12sn7. good 6sn7 now is too expensive.
12sn7 still can use reasonable to buy excellent tubes.
tx
thomas |
|
|
| tube-lover |
hi cbutterworth,
probably the circuit will had little different between your choice & mine.
During I test for the circuit. I had experience use dale resistance in the input tubes plate loading & cathode resistor & generate some noise( because of temp rise) , Pls try to use large watts dale or parallel two samll one Otherwise use Tantulum resistors.
hope this can help.
thx
thomas |
|
|
| burnedfingers |
Hi Charlie.
I did some tube rolling to see if I could come up with noise problems on my 6SN7/6SN7 Aikido.
I did
I have noise if I use any of my Sylvania chrome top green lettered tubes. I also have noise with Tung Sol and GE.
My Aikido is completely quiet if I use my 5692's or my nickle plated Russian equivalents or my cheep Russian tubes in the input position.
The Russian tubes I like the best are the 1578's check ebay# 250089132589 for a picture. Very nice tubes. The standard cheap 6N8S are quiet also.
This is in the input position I am talking about. I haven't tried rolling tubes in the second position yet but will do so later today. |
|
|
| burnedfingers |
Charlie,
Could you post your schematic and power supply? |
|
|
| tube-lover |
hi burnedfingers,
I check on ebay about your advise tubes.
The Russian tubes I like the best are the 1578's check ebay# 250089132589 for a picture. Very nice tubes. The standard cheap 6N8S are quiet also.
price was quite high in russian tube. can U let me know the sound different between NOS tubes, only more silent!!!
I choose 12sn7 because NOS was quite cheap, i buy many two years before to keep stock. I love 12sn7 much more than 6sn7.
thx
thomas |
|
|
| burnedfingers |
Thomas,
I generally like the GE and the Chrome top Sylvania tubes because they have a great midrange and bass response. The 5692's are probably the best american tube in my opinion. The midrange is very smooth and detailed but the tubes are expensive.
The 1578's the nickle base Russian tube is every bit what the 5692's are and more. This tube seems to be even better
in the midrange and high frequency than the 5692 is. The standard Russian 6N8S is a quiet tube but nothing special. I don't feel it has the response the Sylvania or GE has.
The 1578 is a great value for the money. It can be purchased for less than 1/2 of what the 5692 brings. Personally I feel its a much better tube all around.
| quote: | | I choose 12sn7 because NOS was quite cheap, i buy many two years before to keep stock. I love 12sn7 much more than 6sn7 |
Thomas do you find the 12SN7 tube to be more quiet than the 6SN7? |
|
|
| analog_sa |
| quote: | Originally posted by burnedfingers
I generally like the GE and the Chrome top Sylvania tubes because they have a great midrange and bass response. |
Are you able to supply some measurement data confirming the "great midrange and bass response"? |
|
|
| cbutterworth |
My PSU schematic is as follows:
1. 125-0-125 Hammond transformer
2. 1N4007 dioide bridge rectifier, each dioide paralelled with a 0.1uF Wima cap.
3. On the output of the rectifier, I have schottky dioide
4. 10H choke
5. 45uF motor-run cap
6. 10H choke
7. 45uF motor-run cap
I also have a snubber, which did not add any noise. I get around 270VDC from the PSU under load from 4 6SN7s. I did add a 0.68uF Wima to make the filter a cap-input filter. I got around 315VDC this way, but with slightly more hum. When modelled in PSUD, if I increase the 0.68uF cap, I get better ripple filtering, but a much larger voltage. If I reduce the 0.68uF I get a slight voltage drop, but poor ripple filtering (way worse than having choke-input). There is a slight possibility that I am still getting some choke ringing due to too low current draw.
Now that I think about it, my potential divider for referencing my heaters to B+ is a little off. I wanted R1 to be 300K and R2 = 100K, as this would give exactly 1/4 B+. However, I couldn't find a 300K 2W at the time and settled for a 330K 2W. This actually give a heater reference of a little over 23%, not the 1/4 recommended by Broskie and others. Maybe this little difference is resulting in some AC getting into the B+???
Currently, I am working on getting a voltage doubler to bring my 6.3 transformer up to 12.6 V. Then I'll rectify it, so that I can use 12V tubes.
My schematic is as the Broskie Stereo Octal PCB. If you want the exact part values that I am using, please let me know.
Regards.
Charlie |
|
|
| SY |
| quote: | | Maybe this little difference is resulting in some AC getting into the B+??? |
Doubtful. |
|
|
| cbutterworth |
Sy,
Great, in that case I'll not worry about dashing out to hunt for a 300K resistor.
While, your around, could I ask how one goes about calculating how much current each tube is drawing based upon the B+ voltage and cathod resistors, etc. I would like to find out how much power the plates are dissipating.
Charlie |
|
|
| Stixx |
| quote: | | could I ask how one goes about calculating how much current each tube is drawing based upon the B+ voltage and cathod resistors, etc. I would like to find out how much power the plates are dissipating. |
I am not Sy, but I learnt that (being an absolute noob) by reading books (Morgan Jones for example) and by applying Ohm's law. Especially the latter is VERY helpful... |
|
|
| cbutterworth |
Stixx,
I am aware of the Morgan Jones book (I do actually own it). However, if I use Broskie's table of octal triodes, eg a 6SN7 with a cathode resistor (Rk) of 1K, I get 0.3A ! Surely this is not correct, even I can guess that. His table does say, however that lk would be 4.5mA.
I'd appreciate an explanation of how Broskie achieves these figures.
Thanks,
Charlie |
|
|
| SY |
You need to look at the tube curves. Remember, the tube is a stupid hunk of glass and metal and doesn't know where it is in the circuit, only what the relative potentials between its terminals are. So let's say you've got a 300V supply; in the Aikido, that means each tube has 150V across it, right? Would you agree that the likely bias voltages (on the order of 5V) are negligible compared to that 150V? So let's neglect them for the purpose of simplifying the analysis.
For the desired operating current, we look at the 6SN7 datasheet and see what relative grid to cathode voltage we need to get the desired current at the expected plate-to-cathode voltage (in this case, slightly under 150V). With that Vgk in hand, we then use Ohm's Law (just like Stixx said) to calculate the bias resistor size, since we now know the required voltage and current. We'll be off a fraction from neglecting to correct the anode-cathode voltage for the small drop of the cathode bias resistor, but that will be swamped by the actual variation tube-to-tube. |
|
|
| burnedfingers |
Charlie,
If you have questions about your choke you might want to check out this thread.
"How to chose the right choke" by Mikael Abdellah
This thread also has a link to a very nice choke calculator. Look at choke filtered supply.
WWW.hagtech.com/theory.html#choke |
|
|
| burnedfingers |
| quote: | | Remember, the tube is a stupid hunk of glass and metal and doesn't know where it is in the circuit, only what the relative potentials between its terminals are. |
SY,
I only hope the tube gods will forgive you.:bawling: |
|
|
| cbutterworth |
Sheldon,
I visited the Hagatech site and looked at their choke calculations. I have been running the 6SN7s using 1K and 680o at 270VDC. This is much lower than current draw than specified for choke input. I was already aware of this from reading the Morgan Jones book. On recommendation from Bruce Anderson, I did add a 0.68uF cap prior to choke #1. The voltage was higher (325VDC), but hum was still present, so I removed it.
Anyway, after reading your suggestion, I revisited the choke / current issue. In the end, I settled on using the 0.68uF cap, but I also added a 20K bleeder across the final cap in the PSU. This ensures that the current draw is more like 35mA, and results in a voltage of 296VDC (I think, it may be 298VDC, but can't remember). Morgan Jones's calculation for minimum current is VDC/H which gives 296/10 = 29.6mA. OK, so I am now pulling at least enough current. The 20K resistor is rated at 10W, and should only need to dissipate 4.5W. However, the cast resistor gets very hot (at least to the touch), so I am going to replace the one 20K with two paralelled 40K 10W, resulting in 20K resistance with each resistor having to dissipate 2.5W.
Anyway, the hum remains the same, which is very tolerable, but audible.
OK, about the tubes - I am going to post this as a new thread. BUT.... When I use the same tube brand and type for input and output tubes, I get NO HUM. Of course, I do not have four of any one tube. If I put my Sylvania 6SN7GTA in one side - no hum; my Tung-sol 6SN7GTB in the other side - no hum. If I mix the two, I get hum. So, could it be that two tubes of the same brand and type are somehow cancelling the hum???? Is this hum likely to be heater AC hum???
Regards,
Charlie |
|
|
| Bruce Anderson |
Hi Charley
I'm back in town. You have a slight hum and a solid B+ supply. Its unlikely your B+ is causing hum, with or without the .68 cap. (There is reason to increase your voltage, but that can wait for another day). You are using AC heaters. Odds are much greater that is causing minor hum. You can always go to DC (as I do) but I would like you to try one relatively cheap and easy fix first. I used this on my AC heated power amp driver to good effect. Basically it is a somewhat crazy over the top fix (my specialty). You have small low value ceramic caps across the heater pins. Now. take the same type cap and solder one to each heater pin on one tube. Other end of each connected to each other. Take a large, higher voltage cap and connect from that point (where they join) to ground (I use 100uf, 450 volt which is probably much larger than is required). Now, on another tube solder your .01 1K ceramic caps from each heater pin directly to ground. That should make your AC as pure as its going to get without a major project. |
|
|
| Bruce Anderson |
Hi Charley
Sorry, didn't read the last paragraph of your post #19. Sounds like you have totally solved your hum problem with properly balanced tubes. If it was me I still would do my "fix" but sounds like you just need to buy more tubes!!! Lets get together again when everything is sorted out. |
|
|
| cbutterworth |
Bruce,
That is a great idea, although if I did go to 12V for the heaters, I might as well go for rectifying it for a few dollars more. Having seen the price of 12SN7s compared with 6SN7s, I am thinking that I could do a lot with the money that I would save doing this.
The Broskie PCB does allow for jumpering to permit use of 6V heaters with 12V supply.
What did you use to recify your 12V heaters?
Charlie |
|
|
| burnedfingers |
| quote: | | Now that I think about it, my potential divider for referencing my heaters to B+ is a little off. I wanted R1 to be 300K and R2 = 100K, as this would give exactly 1/4 B+. However, I couldn't find a 300K 2W at the time and settled for a 330K 2W. This actually give a heater reference of a little over 23%, not the 1/4 recommended by Broskie and others. Maybe this little difference is resulting in some AC getting into the B+??? |
Charlie,
How much DC is being referenced to the heaters?
Joe |
|
|
| cbutterworth |
I haven't actually measured the voltage, but I have a B+ of around 298 VDC. Resistor 1 is 330K, resistor 2 is 100K, which should result in around 69V ar 23% of B+. Before I changed PSU configuration, I was getting about this percentage.
Charlie |
|
|
| cbutterworth |
Sorry,
I do actually have 300K and 100K, which gives around 72V measured for 292VDC in. This is pretty much 1/4 B+
Now, I have this potential divider located across the final cap in my B+ supply. The 100K has a 0.1uF cap across it.
John Broskie suggested that I try putting the potential divider between B+ and the GND pad on the PCB. He has also suggested that I place two 1K 1% resistors across the transformer winding, and reference the join of the two resistors to 1/4 B+ He has further indicated that DC heaters are much better, and that 12SN7s have better bass resolution.
Bruce, if you read this, I am increasingly thinking that 12VDC heaters may be the best way in the long term, especially as I'll then be able to experiment with 12SN7s and 12SX7s as well as 6SN7s. I suppose that I could try running the heaters from the 12VDC power supply on my Heath PSU, or even a car battery. A lack of hum using these would indicate that my hum is resulting from Ac heaters.
Charlie |
|
|
|
|
