| woodturner-fran |
I don't know what happened to my original post!!
Anyway, no matter how I search for this I can't seem to get a straight answer.
How do I measure hum? Is it as simple as putting a DMM set to mV AC across the loudspeaker terminals with nothing connected to input?
What measurement should I get (I know zero would be great but in the real world whats OK)?
I'm awaiting my morgan jones books!
Fran |
|
|
| gofar99 |
Hi, There are a number of ways to measure hum. I prefer using an oscilloscope. (hope you have access to one) You can use a sensitive meter, but there are two problems. First you only know that you are measuring a voltage, it could be noise, hum, part DC, or in the case of a recent amp of mine high frequency oscillation. The scope tells it all. Second depending on what is actually there the meter might not be accurate (not linear) at the frequency of the measured voltage(s). As for how much is too much, this is a personal factor. Zero is best but rather unusual. Solid state amps are often quieter than tubes. In either case I prefer to have less than 5millivolts of hum going to the speaker. That works out to under a milliwatt and unless you have rather sensitive speakers will not be audible except when you put your ear right up to the speaker.
My present tube amp has about 10-15 mv and when driving 87 db/w speakers is barely audible at the speaker. It is completely gone at anything over 6 inches.
Hope this helps |
|
|
| woodturner-fran |
Thanks,
I put my DMM across the speaker terminals last night and got 3.5mV one side and 5.7mV the other side. From what you're saying this would be low enough (and I don't hear the hum from my loudspeakers) but I can hear it from the headphone jack - which is basically connected into the valve output before the output transformers.
I'm awaiting some better tubes to see if this helps reduce it, and failing that I think my next step would be to convert the AC heater supply to DC.
Fran |
|
|
| SY |
| If I make the unjustified assumption that you've got 8 ohm speakers, 1W is 2.83V. Your hum is 55-60dB below that. That's OK, but 1mV is easily attainable without DC heaters. A scope will help a LOT to tell you where to do the last bit of tweaking. |
|
|
| woodturner-fran |
SY,
you know this is to do with that chinese 6P1 amp that I put up a thread about a little while ago. The new mil-spec tubes haven't arrived yet, but I just wanted to put some figure on the hum I have now, and then again after putting in the new tubes....
An oscilloscope would be handy but I don't know of anyone with one, never mind whether I would be able to figure out exactly what to do with it!
stupid maybe, but sometimes having the figures really helps.
Fran |
|
|
| SY |
| Compared to what one spends on tubes and transformers, a scope is pretty cheap and as necessary to amp design/building/tweaking as socket wrenches are for engine work. Without it, you're working blind. |
|
|
| woodturner-fran |
Can you give me some pointers as to what to look for? The picture I have in my head of an oscilloscope is of this kind of thing:
Wouldn't these be expensive? I would've said 300+ ?
Maybe theres some simpler handheld modern ones or even gizmos that connect to a PC?
Fran |
|
|
| DrExotica |
| Along these lines, I've been meaning to pick up a scope for ages. What are some reasonable recommendations if your budget is $250 or so? Also, are there any good tutorials or other information (books, whatever) that tell you how to effectively use one on tube gear? My only experience with scopes was years ago designing digital circuits. |
|
|
| gofar99 |
All I read above rings true with my experience. Two things stand out first a modest scope can be had on ebay for under $100US. It's a good investment if you intend to continue with audio changes. Second if you are getting the headphone signal off the tube plates, I see a number of problems. First how are you blocking the high voltage DC. This could be dangerous. Second the level of the hum will be higher as the overall voltage of the signal is higher (lower current though). This is why there is a transformer. I would take the signal off the output side of the transformer and use some sort of dividing network to bring it down to reasonable levels. I have seen simple networks using a series resistor of about 50-100 ohms and a second one parallel with the phones of about 8-20 ohms. I use a 50 ohm and 8 on my K-12 and it works fine.
Good listening
gofar99 |
|
|
| SY |
| We've had a PILE of threads on choosing a scope. Figure to spend $100-300 for something that will give you years of good service. |
|
|
| kevinkr |
I highly recommend the Tektronix 2245A which with careful shopping around on eBay can be had for about $300. (That's what I paid for mine last year.)
PC based scope cards and tube audio don't mix, one accident and both card and pc are history. I know someone who did this.
There are a lot of inexpensive early solid state tek scopes on eBay. They're pretty reliable, but in all cases make sure you are buying one from someone who knows how to test a scope and guarantees that it is in full working order. Don't buy a broken one thinking you will fix it, parts are very expensive in most cases if available at all.
Digital sampling scopes blow unless you have big bucks for one of the latest models. (Like a Tek Digital Phosphor.. )
New Kikusui's are an option here and overseas and should sell for $300 or less. Used Philips analog scopes are nice too, stay away from the early digital versions though - too many artifacts in the trace.
Feature set: dual channel, 20MHz vertical bandwidth or better, delayed time base with 50nS/division or better, vertical sensitivity of at least 5mV/division, internal calibrator. Scope probes 1X/10X switchable or a pair of each. |
|
|
| kevinkr |
In terms of hum SY is right, less is better. My current amplifier has hum and noise output of <0.5mVrms and is very quiet with my speaker system approaching 100dB sensitivity.
The scope you posted would be a good choice, and should be available for a lot less than $300. It's quite ancient.. (Late 1980's vintage in all likelihood.)
I use both a scope and a wideband rms meter to do my noise measurements. I also have several different filters including an 'A weighting' filter and a filter that has a bandwidth of 10Hz - 30kHz for unweighted measurements.
|
|
|
| gofar99 |
Less hum is always better. I bought a B&K 30MHZ Dual trace on ebay a year ago for $100. Another feature that I find handy is the built in square wave source. Only 1000HZ in mine, but very handy. The residual hum in my K-12 is likely the result of changing to ultralinear mode on the outputs. It effectively bypasses one section of the ps filter (goes to the screens). I may fix that at another time, but the level of hum I now have is not audible at the listening spot. It's very existance kind of bugs me, but not enough to fix it just now. The heaters are AC, but still I think that 1-2mv level should be possible with this amp with out much trouble.
Good listening
gofar99 |
|
|
| woodturner-fran |
SY - I search here for scope choices. Don't know how deep I'm gonna end up here so I'll have to check around and see whats available.
gofar99 - re the resistor network, is what you mean like a line attenuator where you connect a resistor in series (on signal) and then another resistor (of about 1/10 value of series one) between signal and ground?
That would be easily enough done but my concern is that I've heard of line attenuators killing SQ, particularly dynamics.
Fran |
|
|
| SY |
My preference is 100MHz (although if thatr's the deal-killer, that can be compromised), dual-channel, 1mV sensitivity or better. I have a Sekusui that I bought second hand, refurbed and calibrated, about 15 years ago. A reliable companion.
I would not turn up my nose at a Tek. :D |
|
|
| gofar99 |
I'd love a 100 dual trace too, but $$.
The resistive network, yes sometimes it can cause problems. I haven't personally had it happen though. The network has two general purposes. First to get the signal down to something the phones can handle, roughly 1-10 milliwatts. Second to provide some sort of load to the amp. Some get quite upset without a decent load. The actual values of resistors can be adjusted quite al lot. I have even seen "L" pads used for this function. Elsewhere I have seen networks (of passive parts) that are supposed to restore or enhance the sound going to headphones when they are driven by a power amp. I have never had to resort to such though. One thing that does come to mind is that headphones have more separation than speakers. This already creates an artificial listening condition. Some of the circuits I've seen are also supposed to provide a certain amount of selective crosstalk to fix that problem. Perhaps another member can provide you with a link to such a circuit.
Good Listening
gofar99 |
|
|
| ArtG |
| Morgan Jones' companion book "Building Valve Amplifiers" has a pretty good discussion of the basic use and selection of 'scopes. This material, of course revolves around the use of the instrument for audio purposes, although the techniques described are also fundamental to many other applications. |
|
|
|
