I have this scope. While it's good, DON'T connect it to B+ directly. The 10X setting on the probe is not enough and you will quickly blow it up. If you get a 100X or 1000X probe, you could try, but better would be to AC couple the scope to the PSU with a small (0.01u?) cap instead.
The 12AX7 SRPP inherently generates plenty of distortion and noise which cannot be circumvented. It has insufficient current carrying capacity to perform the task and suffers from instability, very high gain, poor bandwidth, frequent saturation events, unfavorable spectral characteristics...
Ok, so just to connect oscilloscope to GND and B+ via small high voltage capacitor (eg 10 nF / 600 V)?
What should be acceptable AC voltage level at B+ to avoid audible hum / buzz?
If it is too high, is the right thing to do to increase capacitors or add a choke?
Maybe I can measure the AC ripple on B+ using a multimeter to have an idea how much is it?
What should be acceptable AC voltage level at B+ to avoid audible hum / buzz?
If it is too high, is the right thing to do to increase capacitors or add a choke?
Maybe I can measure the AC ripple on B+ using a multimeter to have an idea how much is it?
Yes. Run a small cap from power to ground with a 1M resistor, and connect the probe to the cap/resistor junction. The lower the noise, the better but I've made power stages that had 1V of ripple that were silent. It all depends on how well the CMRR of the OPT is really.
When you connect a cap from B+ to the scope probe, during the charging time of that 0.001 or 0.1 uF cap, All of the B+ will be at the probe tip. That can damage the scope channel input.
Build your own B+ attenuator. Use a 10 Meg series resistor, and a 110k resistor to ground. The 110k resistor in parallel with the scope channel's 1 meg input resistor is 100k to ground.
That will divide the B+ by 100:1. You do not have to use frequency compensating capacitors, you are looking at hum at 1x 2x and 3x mains frequency.
50, 100, 150; or 60, 120, 180 Hz.
Also, with a 500V B+, and 100:1 divider (+5V out of the divider), you can safely use the scopes AC coupling, and see the ripple with the scope sensitivity set at 20mV, 10mV, 5mV, and 1mV per division. Then set the channel to 100x, to read out the true ripple voltage.
That will be 2V, 1V, and 0.5V per division. Set the trigger source to 'Line'. With the trigger set to line, you can set the scope acquisition to average, and set the average count to a high number. That way you can average out line bounce, and high frequency noise, so that you can see the ripple.
I use a professional 100 x scope probe (divides by 100). But you can build your own as per the above.
If you can not see the ripple because of the 100:1 division, then use a 10 Meg Ohm series resistor, and a 1 Meg Ohm resistor to ground. That 1 Meg in parallel with the scope 1 Meg is 0.5 Meg. That forms a 20:1 divider.
Do the same as you set up for the 100:1 measurement.
Do not let your lack of professional test and measurement equipment stand in your way of making a good measurement.
I had to do a lot of special measurements and setups when I was in the middle of the Pacific, but I did not have the test equipment and parts I would have wanted. The captain says "Fix it". You get the job done.
Build your own B+ attenuator. Use a 10 Meg series resistor, and a 110k resistor to ground. The 110k resistor in parallel with the scope channel's 1 meg input resistor is 100k to ground.
That will divide the B+ by 100:1. You do not have to use frequency compensating capacitors, you are looking at hum at 1x 2x and 3x mains frequency.
50, 100, 150; or 60, 120, 180 Hz.
Also, with a 500V B+, and 100:1 divider (+5V out of the divider), you can safely use the scopes AC coupling, and see the ripple with the scope sensitivity set at 20mV, 10mV, 5mV, and 1mV per division. Then set the channel to 100x, to read out the true ripple voltage.
That will be 2V, 1V, and 0.5V per division. Set the trigger source to 'Line'. With the trigger set to line, you can set the scope acquisition to average, and set the average count to a high number. That way you can average out line bounce, and high frequency noise, so that you can see the ripple.
I use a professional 100 x scope probe (divides by 100). But you can build your own as per the above.
If you can not see the ripple because of the 100:1 division, then use a 10 Meg Ohm series resistor, and a 1 Meg Ohm resistor to ground. That 1 Meg in parallel with the scope 1 Meg is 0.5 Meg. That forms a 20:1 divider.
Do the same as you set up for the 100:1 measurement.
Do not let your lack of professional test and measurement equipment stand in your way of making a good measurement.
I had to do a lot of special measurements and setups when I was in the middle of the Pacific, but I did not have the test equipment and parts I would have wanted. The captain says "Fix it". You get the job done.
So have made some progress in analyzing the problem - it seems that adding choke to B+ Power supply CLC filter helps - very simply measuring spectrum of noise coming out of the speaker using iPhone shows the peak at 100 Hz. If I add the choke the peak goes down by ~ 8 dB. Is there anything else to improve it further? More capacitors in CLC filter?
As I mentioned before there is also an audible noise coming from toroidal power transformer (50 Hz). I think I will try to use to use some filter to eliminate DC voltage at on transformer mains.
One more thing I want to try is to use DC voltage for tube heaters - both 12AX7 as well as 6L6GC. One question here - I will use a rectifier and LT1085 regulator - should I leave both outputs (plus , minus) connected just to heaters floating or rather ground the minus?
As I mentioned before there is also an audible noise coming from toroidal power transformer (50 Hz). I think I will try to use to use some filter to eliminate DC voltage at on transformer mains.
One more thing I want to try is to use DC voltage for tube heaters - both 12AX7 as well as 6L6GC. One question here - I will use a rectifier and LT1085 regulator - should I leave both outputs (plus , minus) connected just to heaters floating or rather ground the minus?
Heater should be referenced to ground in some way. I usually tie - to ground, but you could tie + to ground if you need a negative rail (for a CCS or something).
I don't think so you will improve using a DC voltage, if there is a problem you get it with DC or AC, I made so many test in this regards and now for my project I use AC voltage for the heaters without any problem....all the best
So I spent some time experimenting and what I found is that no real benefit of using DC filament voltage, so I switched back to AC with center tap elevated to a resistor divider between B+ and GND.
Using a choke instead of resistor definitely helped (CLC) for B+
What I’m finding now is:
- when I open the top cover where output transformer are mounted I have almost silent output from speakers - my iPhone mic spectrum analyzer shows something like -80 dB
When I close the top cover or move it closer to the chassis I can see a peak of - -69 dB / 150 Hz for the left channel and -70 dB / 50 Hz for the right channel —> hum goas up by ~ 10 dB and becomes audible - different hum from left (150 Hz) and right (50 Hz) channels
- by closing top cover (aluminum chassis) I’m moving top mounted output transformers closer to a choke (left channel) and main power toroidal transformer mounted in the main chassis
So it seems the hum is somehow caused by coupling between output transformers and power transformer, choke,....
Also when I remove tubes from one channel the hum peak drops down by ~ 5 dB (happens with both left and right channel).
No effect of removing input tunes only, when I remove all tubes, amp is almost silent.
All chassis is made of aluminum. What I’m thinking of as a next step is to cover main power transformer by a steel cover and also add steel covers to output transformer windings which are open now and heading towards main power transformer ajd choke when top cover closed.
Using a choke instead of resistor definitely helped (CLC) for B+
What I’m finding now is:
- when I open the top cover where output transformer are mounted I have almost silent output from speakers - my iPhone mic spectrum analyzer shows something like -80 dB
When I close the top cover or move it closer to the chassis I can see a peak of - -69 dB / 150 Hz for the left channel and -70 dB / 50 Hz for the right channel —> hum goas up by ~ 10 dB and becomes audible - different hum from left (150 Hz) and right (50 Hz) channels
- by closing top cover (aluminum chassis) I’m moving top mounted output transformers closer to a choke (left channel) and main power toroidal transformer mounted in the main chassis
So it seems the hum is somehow caused by coupling between output transformers and power transformer, choke,....
Also when I remove tubes from one channel the hum peak drops down by ~ 5 dB (happens with both left and right channel).
No effect of removing input tunes only, when I remove all tubes, amp is almost silent.
All chassis is made of aluminum. What I’m thinking of as a next step is to cover main power transformer by a steel cover and also add steel covers to output transformer windings which are open now and heading towards main power transformer ajd choke when top cover closed.
well the first thing I would do is take away the power transformer from the chassis and put it far away with extension cables and see what is the result.
If you are using a toroidal power transformer then I'd suggest you firstly confirm there is no inadvertent dc bias causing BH loop asymmetry such that one side of loop is entering saturation region. The mechanical humming of the PT may be an indicator.
If there is BH saturation occurring then I'd suggest you eliminate that first before chasing other ways to minimise hum, as you may be seeing smoke and mirrors.
If there is BH saturation occurring then I'd suggest you eliminate that first before chasing other ways to minimise hum, as you may be seeing smoke and mirrors.
Typical Hum Level in 300B SET?
Hope I'm not interrupting here, but I am interested in learning what hum level is typical in 300B SETs. I have a Audio Electronic Supply SE-1 (a Cary design) and I'm getting between 1-2 mV of hum at the outputs into 4 ohm speakers (or an 8 ohm dummy load) with the inputs shorted. This is almost inaudible at the listening position with my ~86dB sensitive speakers, but I'm considering building a high-efficiency single driver speaker and would like to reduce the hum level of the SE-1, if possible.
Any thoughts or benchmarks would be most helpful.
Hope I'm not interrupting here, but I am interested in learning what hum level is typical in 300B SETs. I have a Audio Electronic Supply SE-1 (a Cary design) and I'm getting between 1-2 mV of hum at the outputs into 4 ohm speakers (or an 8 ohm dummy load) with the inputs shorted. This is almost inaudible at the listening position with my ~86dB sensitive speakers, but I'm considering building a high-efficiency single driver speaker and would like to reduce the hum level of the SE-1, if possible.
Any thoughts or benchmarks would be most helpful.
No success so far 🙁
Well I have done some major changes like replacing toroidal power transformer by EI type (no mechanical noise now!), moved power transformer outside of chassis and oriented it 90 desgrees to output transformers. Also added 30% more filter capacitors and did some wiring clean up.
What I’m seeing now is:
- amost no effect of power choke (tried to replace choke with resistor)
- there is a hum of 150 Hz in both channels... (why 150 HZ and no 50 Hz?)
- it is audible even when I disconnect all input and output tubes
- I tried to shortcut input, disconnect heating 6.3 V as well as B+ power voltage 260 V and no change....
- the only audible change happens if I switch output from 4 ohm to 8 ohm (little bit less hum) ot switch headphones instead of speaker output (resistos divider) - in this case it is completely silent in speakers
What can be wrong if there is a hum even without tubes and with B+ discunnected?
Well I have done some major changes like replacing toroidal power transformer by EI type (no mechanical noise now!), moved power transformer outside of chassis and oriented it 90 desgrees to output transformers. Also added 30% more filter capacitors and did some wiring clean up.
What I’m seeing now is:
- amost no effect of power choke (tried to replace choke with resistor)
- there is a hum of 150 Hz in both channels... (why 150 HZ and no 50 Hz?)
- it is audible even when I disconnect all input and output tubes
- I tried to shortcut input, disconnect heating 6.3 V as well as B+ power voltage 260 V and no change....
- the only audible change happens if I switch output from 4 ohm to 8 ohm (little bit less hum) ot switch headphones instead of speaker output (resistos divider) - in this case it is completely silent in speakers
What can be wrong if there is a hum even without tubes and with B+ discunnected?
I'm not an expert but I had a similar problem and it was due to grounding issues, specifically that my headphone jack was grounded directly to the chassis and messed up my star grounding layout.
You're probably familiar with star grounding but here is a good reference: http://www.geofex.com/Article_Folders/stargnd/stargnd.htm
You're probably familiar with star grounding but here is a good reference: http://www.geofex.com/Article_Folders/stargnd/stargnd.htm
magnetic coupling, had it happen to me and it drove me nuts, took physical relocation and some mu-metal shielding to reduce (not eliminate) it so it was no longer objectionable.
If there is hum when the output tube(s) are removed you have some magnetic
field in your room.
Any power lines nearby ? Any large transformers in the building ? Can
you move the amp to another room and is there any change in hum-level then ?
Hi
Have you tried to twist all the AC cables? Yet those of small voltage, like the 6Vac of filaments?
Have you tried to twist all the AC cables? Yet those of small voltage, like the 6Vac of filaments?
It seems it is really coupling between power and output transformers. My chassis is pretty small and transformers are very close to each other. I played a bit with them and when I rotate power transformer by 90degrees hum dropped by -15 dB - great! Now the hum is audible only with ear ~10 centimeter close to the speaker. Is it normal or should I still try to optimize it somehow? (It is SE using 6L6GC, zero feedback, speaker 98 dB sensitive horns).
Now when I plug tubes to sockets and power it on I can also hear some noise (like white noise) in both channels (I think it is ok) and one channel is also “buzzing” (like bzzzz). One channel only. Any idea what can cause buzz in one channel only?
Now when I plug tubes to sockets and power it on I can also hear some noise (like white noise) in both channels (I think it is ok) and one channel is also “buzzing” (like bzzzz). One channel only. Any idea what can cause buzz in one channel only?