I would like to move forward and get my own test equipment so I can learn more and not have to rely on others. Is there a unit that would combine both the signal generator and scope for easier cabling and simplicity? My budget would be up to about $150 and hopefully I could get it from Amazon.
-
Please note that there are a few updates and clarifications made in the Audiokarma Rules, mostly relating to advertising and the addition of the new "Paying it Forward" & "Giving back" forums in the AudioKarma Audio Marketplace section.
El34 / 12AX7 SE Build (My 1st build)
- Thread starter John Berard
- Start date
I highly recommend getting a used Analog Discovery 2 off ebay, and then download the freeware Audio Analyzer Suite.
wparks
Super Member
Most of the inexpensive scopes do have a rudimentary signal generator of some sort built into them, but I would not recommend using that as your primary source. The generator built into mine, anyway, does not have many features, and does not even allow setting the amplitude.
Let's back up a second- The Analog Discovery device and software that Skunkie uses is very powerful for amplifier characterization. It is essentially an analog to digital converter (ADC) to digitize incoming (amp output) signals, and a DAC for synthesizing outgoing (amp input) signals. The output signals are generated by the software, and the input signals are visualized and analyzed by the software that can do very sophisticated analysis. It is most well suited to amplifier characterization- being very easily able to measure gain, power output, frequency bandwidth, and most importantly, distortion. Measurements across sweeps of frequency, and distortion, are difficult or impossible to do with just a signal generator and oscilloscope. So, if you are jumping in with both feet, definitely consider the Analog Discovery. If you are on a budget, like me, and want to get more nitty-gritty, you can build your own version of what the Audio Discovery is doing by yourself. This route would be using a decent USB "sound card" type audio interface card/box combined with the laptop and free software, and adding a more flexible/adjustable/powerful op-amp based input/output hardware front end to the sound interface. This is detailed in Rod Elliott's ESP project #232 "Distortion Measurement System". I'm on a budget, and do a lot of building with op-amps so this is very high on my "next project list".
An oscilloscope is a very handy piece of general debug and test equipment that every workbench should have for many jobs above and beyond amplifiers. You will end up using the scope to look at power supply noise/ripple, power up transients, evaluating signal presence/amplitude/quality from stage to stage and many others. I use them to characterize transformers- injecting an AC signal with my generator into the primary, and measuring input vs output amplitude to determine power transformer and audio transformer turns ratios (to calculate voltage and impedance ratios). With a non-vertically symmetrical waveform (like a ramp) you can even determine winding polarity, (to use a secondary winding to buck the primary of a power transformer down so the filaments are at the correct voltage), determining polarity of an output transformer, or even to verify input to output polarity of an open-loop amplifier before you start connecting and tuning feedback. Then a scope is handy for analyzing amplifier transient performance and overall stability, using 10kHz square wave behavior, such as ringing, output transformer resonance, or detecting amplifier oscillation with capacitive loads, etc.
It would serve you well to do some reading and research here, and get an O-scope that meets your needs. In general, I recommend a 2 channel unit, anything able to handle into the 10's of megahertz or above is fine, rechargeable/battery powered for isolation and safety, and with a collection of different probe options. Sometimes you will want a high voltage 1:1 or 10:1 probe, sometimes alligator clip leads, and sometimes just an RCA to BNC adapter (shown below). For audio work you CAN do with a scope, any cheap RCA cable has more than enough bandwidth. Using RCA patch cables from the signal generator to the amp, and a splitter to go into the scope input, etc, makes this work really quick and easy. I added some RCA ports to my speaker load, and a simple RCA cable from the load to the scope makes it really simple. A couple of single male to double female RCA splitters (shown below) allows me to inject a mono signal identically into both sides of my amplifier, and another branches a version of the input signal for measurement by the scope- so while doing gain (output/input) measurements I can easily move the scope back and forth to measure the output and input amplitudes. Overall, it's not very much cabling (my typical setup also shown below.)
For a signal generator, most these days are microprocessor based instead of analog oscillators. This is great because you can set it to a precise frequency (like 1Hz, or 10,001 Hz) and it will be correct. Because the output waveform is synthesized numerically out of a DAC, the waveform will have precisely the same amplitude at any output frequency you set, which is great for spot checking bandwidth. (evaluating output amplitude at a couple of different frequencies from very low to very high) without having to constantly re-adjust the input amplitude to be the same for each frequency. You will want to be able to dial the output signal amplitude from very low (a few mV) to pretty high (5V at least) , with hopefully either a large knob or multi-turn pot to be able to dial the tweak the amplitude to just the right level. Having a host of different output signals is a must, and having an adjustable DC offset is nice, but not often required. My cheapie little $30 unit has served well, but I wish it had a larger knob for setting the output amplitude (instead of the tiny one on the side), and it would be nice to be able to change the output settings without stopping and re-starting the waveform.
wparks
Super Member
BTW- Using just a simple signal generator, dummy load, and basic oscilloscope (my ~$75 setup), you can accomplish all of the following audio performance quality tests:
With Sine Wave:
Check for outright oscillation at first power-on or at no-load
Maximum output power at clipping, vs load resistance
Input signal amplitude for maximum output power (amplifier sensitivity)
Clipping voltage, behavior and asymmetry, Sine appearance near clipping
Output impedance and amount of power supply sag with both channels fully loaded and during transients
Open loop gain, Closed loop gain (measure or set amount of negative feedback)
Check gross channel separation (cross-talk) and amplitude matching
Frequency and power bandwidth vs frequency
Find resonances in the audio path, output transformer resonances
Observe cathode, screen, pre-amp and supply behaviors
With 1kHz/10kHz Square Wave:
Output rise/fall edge rate and symmetry
Ramping at low frequencies (coupling caps)
Overshoot damping/ringing (transient stability)
Overall stability and margin vs. capacitive loading
Feedback optimization for best tonality and gain
Feedback compensation for best frequency bandwidth (flatness) and stability
I cover most of these in my Fat Daddio SE build thread.
With Sine Wave:
Check for outright oscillation at first power-on or at no-load
Maximum output power at clipping, vs load resistance
Input signal amplitude for maximum output power (amplifier sensitivity)
Clipping voltage, behavior and asymmetry, Sine appearance near clipping
Output impedance and amount of power supply sag with both channels fully loaded and during transients
Open loop gain, Closed loop gain (measure or set amount of negative feedback)
Check gross channel separation (cross-talk) and amplitude matching
Frequency and power bandwidth vs frequency
Find resonances in the audio path, output transformer resonances
Observe cathode, screen, pre-amp and supply behaviors
With 1kHz/10kHz Square Wave:
Output rise/fall edge rate and symmetry
Ramping at low frequencies (coupling caps)
Overshoot damping/ringing (transient stability)
Overall stability and margin vs. capacitive loading
Feedback optimization for best tonality and gain
Feedback compensation for best frequency bandwidth (flatness) and stability
I cover most of these in my Fat Daddio SE build thread.
So much to consider! I like the compact size of the units you propose and the Analog Discovery. The Discovery however has the added advanges of being a single box and the ability to measure distortion and harmonics. With the added BNC adapter, a better Oscilliscope solution could be coupled. The Discovery is more $$ overall, but not excessive especially with the added capabilities geared toward my projects. I will sit on the bence for a bit and ponder all the options.
Thanks for your detailed and extensive input on these matters Mr. Parks.
Although not needed for a normal DIY hobbyist, if anybody is into measuring your devices as much as building (like me), check out the Audio Precision ATS-1 test set. It is a standalone unit so it is not limited by a computer soundcard and gives laboratory type measurements which are very accurate. It is an older system so it does have limitations, but you can't beat it for low level resolution.
There are quite a few on the used market. Just FYI.
There are quite a few on the used market. Just FYI.
I have been critical listening to the amp rolling many tubes. My favorites are the Mullard xf2 ver. from the 1960's, follow by some early Svetlana from the 1980s. The JJs aren't to my taste and were a bit too bright for my taste but they did have detail. All the driver tubes are well matched on my Maxi Matcher. I am still waiting to get this on some test gear and I may spring for a Analog Discovery 3.
In the mean time, I went ahead and wired up the UL/Triode switch. It was a bit of a challenge considering the UL and triode mode each have their own feedback. I also added the 100Ω resistors to the 6.3v heater supply.
Now I will be comparing UL to triode mode. I have to keep in mind not to switch between the two while powered on. I had considered a locking toggle as a way of thinking twice before switching but those buggers are expensive. I am sure I will be doing some tweeking of the shade resistors and perhaps other things depending how it all tests on a scope and AD3.
In the mean time, I went ahead and wired up the UL/Triode switch. It was a bit of a challenge considering the UL and triode mode each have their own feedback. I also added the 100Ω resistors to the 6.3v heater supply.
Now I will be comparing UL to triode mode. I have to keep in mind not to switch between the two while powered on. I had considered a locking toggle as a way of thinking twice before switching but those buggers are expensive. I am sure I will be doing some tweeking of the shade resistors and perhaps other things depending how it all tests on a scope and AD3.
Need help guys. This is driving me crazy(er). I wired up the UL/Triode switch circuit and when switched to UL, it plays great. However, when switched to Triode, nothing, nada from both channels. I double checked my connections and didn't find any problems. This leads me to believe that the schematic I designed is faulty. Any help is appreciated!
wparks
Super Member
I understand in triode mode you connect the pentode screen to the plate, that looks correct.
Do you have the "optional 750K", or when the switch is in triode mode are you just shorting the pentode plate to the triode plate as the schematic seems to show? That would be bad. Also, check your 750K value if it is there. If the optional 750K is not desired, that connection should be open so there is no negative feedback back to the first triode.
Do you have the "optional 750K", or when the switch is in triode mode are you just shorting the pentode plate to the triode plate as the schematic seems to show? That would be bad. Also, check your 750K value if it is there. If the optional 750K is not desired, that connection should be open so there is no negative feedback back to the first triode.
Need help guys. This is driving me crazy (er). I wired up the UL/Triode switch circuit and when switch to UL, it plays great. However, when switch to Triode, nothing, nada from both channels. I double checked my connections and didn't find any problems. This leads me to believe that the schematic I designed is faulty. Any help is appreciated!
View attachment 3603254
View attachment 3603254
I have no feedback resistor in triode mode so I guess it's shorted. But it is my understanding that is a zero feedback circuit?
I put in the 750k and everything is working fine now. My 'newbieness' is on full display! Apparently I didn't distroy anthing in the process. Now that I studied how a shade feedback circuit works, I understand it needs resistance to work and would thus need to be open for zero feedback. So much to learn and so few years left.
Once again Mr Parks, I bow to you and thank you!
I will definately double back and check 1.8V biasing measurement is still okay. I probably did shorten the life of the input tube.
Feels good to solve a problem don't it.I put in the 750k and everything is working fine now. My 'newbieness' is on full display! Apparently I didn't distroy anthing in the process. Now that I studied how a shade feedback circuit works, I understand it needs resistance to work and would thus need to be open for zero feedback. So much to learn and so few years left.
Once again Mr Parks, I bow to you and thank you!
wparks
Super Member
No sweat- an easy mistake to make, I'm just glad it was an easy fix.
I am curious to see what your listening impressions are between the two modes. How much of a sensitivity difference are you hearing? I would expect open loop to be louder for the same signal, but switching to triode mode could have less gain so curious how that ends up. You might also notice a bigger difference in the open-loop-triode-mode sound as you crank the volume up- as you approach maximum power the distortion increases significantly, and can, from what I have heard and experienced running pentode amps open-loop, give a sort of "bloated" sound. I'm only building my 5th amp, all pentodes, and I have never built an amp output stage with either ultra-linear, or with schade feedback, or even a triode strapped or native, so I'm interested in learning vicariously through your results. We would love a full report!
I am curious to see what your listening impressions are between the two modes. How much of a sensitivity difference are you hearing? I would expect open loop to be louder for the same signal, but switching to triode mode could have less gain so curious how that ends up. You might also notice a bigger difference in the open-loop-triode-mode sound as you crank the volume up- as you approach maximum power the distortion increases significantly, and can, from what I have heard and experienced running pentode amps open-loop, give a sort of "bloated" sound. I'm only building my 5th amp, all pentodes, and I have never built an amp output stage with either ultra-linear, or with schade feedback, or even a triode strapped or native, so I'm interested in learning vicariously through your results. We would love a full report!
Oh yeah. Hopefully it didn't cost me any $$. Actually it cost me $25. Note I am now a subscriber! Well worth it.Feels good to solve a problem don't it.
I will definately post my impressions. Right now I am checking for any damage I may caused. I put in a fresh input tube and took some measurements. Hopefully the voltages are showing good health? With the mention of possibly hurting the LED, I am wondering if the measured 1.73V is close enough to the spec'd 1.8V?
