I am planing to replace my Parasound HCA855A with some low watt choke loaded amp for driving a set of Mid and High frequency horns.
The reason being that the power needed is less than a watt and the Parasound seem to be noisy ,plus it's a 85W/ch amp. The noise is somewhat less than a millivolt, but I can hear it on the midrange. I will use the inexpensive Hammond 159ZC 60mH 2A choke and use ca. 15V with it. This way I expect ca. 3W clean output by using not more than 30W/CH. To start I was to measure a few different output transistor candidates. I'm using an old HP analyzer ,which is able to measure down to -90dB and using for the signal source either a test CD or an old Bruel & Kjaer beat frequency oscillator's 600 Ohm's output. Here to follow a set of pictures for 1W output with different transistors in a source follower configuration.
The reason being that the power needed is less than a watt and the Parasound seem to be noisy ,plus it's a 85W/ch amp. The noise is somewhat less than a millivolt, but I can hear it on the midrange. I will use the inexpensive Hammond 159ZC 60mH 2A choke and use ca. 15V with it. This way I expect ca. 3W clean output by using not more than 30W/CH. To start I was to measure a few different output transistor candidates. I'm using an old HP analyzer ,which is able to measure down to -90dB and using for the signal source either a test CD or an old Bruel & Kjaer beat frequency oscillator's 600 Ohm's output. Here to follow a set of pictures for 1W output with different transistors in a source follower configuration.
Attachments
The following transistors I tested:
Hitachi 2SJ50
Hitachi 2SK135
IRF240
IRF520
SemiSouth E120R100
Sony 2SJ28
Sony 2SK60
Sony 2SK82
Hitachi 2SJ50
Hitachi 2SK135
IRF240
IRF520
SemiSouth E120R100
Sony 2SJ28
Sony 2SK60
Sony 2SK82
Attachments
Here's it. I have also made a test tone CDR and played back the 1kHz track on a Philips CD921 player, as a low distortion source. Here's the spectrum of that ,the signal going through the LM4562 opamp and it's output at 3V. To start with I will use this IC to drive the output stage, so the last picture is the Semisouth E120R100 at 1W , 1kHz into 8 Ohms with the CDR source.
Attachments
Miklos
Thank you for taking the time to post the source spectrum.
I had no idea that a Bruel & Kjaer beat frequency oscillator could get anywhere near these low levels. Sorry have to ask, are you sure those are 10db steps ?
Interesting stuff, thanks for the real data.
-Antonio
Thank you for taking the time to post the source spectrum.
I had no idea that a Bruel & Kjaer beat frequency oscillator could get anywhere near these low levels. Sorry have to ask, are you sure those are 10db steps ?
Interesting stuff, thanks for the real data.
-Antonio
The oscillator is old and pretty unreliable, but very convenient to use it on the bench.
What is confusing with it that the "Attenuator" output using the generators transformer 60 Ohms tap. For the 1W measurements I needed 3 V output and I got it on the 600 Ohm tap, from the "Load" connector, which can produce 40V, so less drive was needed from the gen. output stage, consequently has less distortion to. Anyhow the measurements are good for comparison of the transistors, fortunately the transistors having higher distortion than the gen. Also I can measure the better ones with the CD+LM4562 setup, but it's already clear at the very expensive Semisouth fet is the best, followed by the pretty inexpensive IRF240.
What is confusing with it that the "Attenuator" output using the generators transformer 60 Ohms tap. For the 1W measurements I needed 3 V output and I got it on the 600 Ohm tap, from the "Load" connector, which can produce 40V, so less drive was needed from the gen. output stage, consequently has less distortion to. Anyhow the measurements are good for comparison of the transistors, fortunately the transistors having higher distortion than the gen. Also I can measure the better ones with the CD+LM4562 setup, but it's already clear at the very expensive Semisouth fet is the best, followed by the pretty inexpensive IRF240.
Choke loaded amp.
It was some time ago when I started this exercise. My goal was to make a good quality easy build single ended 10W/Ch amplifier (gave up the idea to replace the 5 Ch Parasound amp) without using output tubes and expensive transformers.
I like to mention I'm an amateur Diyer with experience, but not EE.
I bought the output transistors, the SemiSouth E120R100 and made four different drivers to try out (criteria was to able to deliver ca. 10V undistorted signal).
The following drivers I built: ECL84 mu stage (I used this before), Sy's His Masters Noise amps second half, a JC2 clone kit from Ebay and a LM4562 IC one. Made a test jig amp and tried all four, measured and played them.
The differences in quality are minor. The tube ones are more difficult to make ,because of the extra two PS, the filament and plate supply, but more fun to have them. Last sprig I finished one complete amp, and I used the HMN driver in it. It was the most difficult driver to make, so might as well use it. The result is very satisfying, sounds good and the power consumption is under 100 W. I'm using them with small two way bookshelf speakers and a powered sub. I used the simple Hammond 159 CZ choke in the amp. These are rated 60 mH 2000 mA and I'm running them ca. 1.9 A. The main high current PS are two HP laptop supply's ,18.5 V 6.5 A 120 W (one could carry the amp). They are glued in with silicon rubber at the back plate of the amp and they are getting just nice warm (like the rest the
amp). I used capacitor multipliers for the outputs and for the plate supply. Just for fun also tested in the HMN a bunch of different dual triode's, without making any mods to the biasing, just plug them in one by one. The HMN is pretty tolerant circuit. Tried the 6N23P, 6N1P, 6GC7 RCA, 8CG7 Germany, ECC88 Holland, ECC88 Tungsram, E88CC Siemens, PCC88 Tungsram, ECC88 labeled AEG,E80CC
Valvo, and ended up using a pair of Tungsram PCC88's, filaments in series. I will try post a few pictures.
ECL84 Mu, HMN, LM4562 and the JC2 clone, in the jig.
It was some time ago when I started this exercise. My goal was to make a good quality easy build single ended 10W/Ch amplifier (gave up the idea to replace the 5 Ch Parasound amp) without using output tubes and expensive transformers.
I like to mention I'm an amateur Diyer with experience, but not EE.
I bought the output transistors, the SemiSouth E120R100 and made four different drivers to try out (criteria was to able to deliver ca. 10V undistorted signal).
The following drivers I built: ECL84 mu stage (I used this before), Sy's His Masters Noise amps second half, a JC2 clone kit from Ebay and a LM4562 IC one. Made a test jig amp and tried all four, measured and played them.
The differences in quality are minor. The tube ones are more difficult to make ,because of the extra two PS, the filament and plate supply, but more fun to have them. Last sprig I finished one complete amp, and I used the HMN driver in it. It was the most difficult driver to make, so might as well use it. The result is very satisfying, sounds good and the power consumption is under 100 W. I'm using them with small two way bookshelf speakers and a powered sub. I used the simple Hammond 159 CZ choke in the amp. These are rated 60 mH 2000 mA and I'm running them ca. 1.9 A. The main high current PS are two HP laptop supply's ,18.5 V 6.5 A 120 W (one could carry the amp). They are glued in with silicon rubber at the back plate of the amp and they are getting just nice warm (like the rest the
amp). I used capacitor multipliers for the outputs and for the plate supply. Just for fun also tested in the HMN a bunch of different dual triode's, without making any mods to the biasing, just plug them in one by one. The HMN is pretty tolerant circuit. Tried the 6N23P, 6N1P, 6GC7 RCA, 8CG7 Germany, ECC88 Holland, ECC88 Tungsram, E88CC Siemens, PCC88 Tungsram, ECC88 labeled AEG,E80CC
Valvo, and ended up using a pair of Tungsram PCC88's, filaments in series. I will try post a few pictures.
ECL84 Mu, HMN, LM4562 and the JC2 clone, in the jig.
Attachments
Distortions at 1 W output (1 kHz 8 Ohm's), but for example with the JC2 clone at 10 W 1 kHz in to 8 Ohm's the THD is 0.64%.
Pictures: ECL84, HMN, LM4562 and JC2 clone.
Pictures: ECL84, HMN, LM4562 and JC2 clone.
Attachments
Some square wave responses , the SemiSouth with the HMN and with the LM4562.
First with the HMN 20kHz, 100kHz 3 V p-p and than with the LM4562 the same, 5 V p-p into 8 Ohm's.
First with the HMN 20kHz, 100kHz 3 V p-p and than with the LM4562 the same, 5 V p-p into 8 Ohm's.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Solid State
- Testing output transistors.