Advanced Audio forum is down

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kingkorg

Well-known member
Joined
Apr 15, 2017
Messages
2,972
Location
Norway
I just realized Advanced Audio's forum is down, and most of the pages can't be accessed even through the Wayback Machine. That forum was a source of some valuable information, as Dave Thomas is very open person and willing to share his knowledge and findings. This way i'd also like to thank Dave for all the info he selflessly shared.

So i managed to dig up this post at least on moding t.bone sct800. Not very likely, but someone might want to go from plate follower to cathode follower.

There was also a quite useful thread on moding Nady 1050, but that one is gone now.


Yes, the SCT-800 was originally imported into North American and Europe from Feilo Microphones in China. These microphones came with an economy 32mm K67 type capsule with more rise at 3-5khz and less rise at 10khz than our AK12.

The circuit works best "as is" with a 12AT7 tube and good transformer with a 10:1 ratio. If you use a 6072a the circuit becomes very similar to the ELA M251 circuit and it requires an output transformer with a 14:1 ratio.

The stock circuit requires a cathode bypass capacitor so you should fit it with a good 200 ufd electrolytic bypassed with a .01 metal film or poly cap which will improve things a bit.

The SCT 800 is a cardiod only but by adding a 9 pattern power supply and in the microphone a 51 meg blocking resistor, a .01 bypass capacitor can have 9 patterns.
The 51 meg is soldered to pin 3 of the 7 pin connector and to the rear diaphragm with the .01 bypassing the back diaphragm to ground as far as audio is concerned.

The SCT-800 circuit and the ELA M251 circuit are nearly identical for all practical purposes. The capacitor across 1 and 2 (1 being the plate and 2 being the grid) increases the "Miller effect" which states, as the internal capacitance of a tube between its signal grid and plate increase the more the High Frequency response will roll-off. The capacitance between the grid and the plate whether internal or external is multiplied by the gain of the circuit. The ELA M251 accomplished this by putting a larger capacitor across the output transformer to tame the HF rise.

The gain of at 12AT7 is 60 so the 1.5pf in parallel with the internal 1.8 pf X 60 = 200pf which will start to roll out the high end response at 10khz to compensate for the rise in the CK12 and AK12's rise in response above 10khz.

In the C12 AKG did not reduce the natural rise of the capsule but they also used fixed bias which lowers the plate output resistance by 1/2.

If you use our AK12 capsule and leave the capacitor in you will get a more C251 like curve or if you remove the capacitor it will have a more C12 type curve.

In the ELA M251 SCT800 circuit with it the de-emphasis capacitor across pins 1&2 the HF response will vary depending on the gain of the tube used. More with a 12AT7 less with a 6072a and even more with a 12AX7.

In our circuit we fit the capacitor between the plate of the first stage and grid of the 2nd stage which offers more buffering. It is not effected by different tube gains in this position.

I believe John Peluso used the same body and ELA M251 circuit with a 6072a tube in his original 22 251 type microphone. He also had a capacitor between 1 and 2 in his 22 251.


The 2 stage CCDA circuit that we use in our microphones will work with a 12AT7 or 6072a which was the tube used in the original C12 and ELA M251. The CCDA circuit provides and even lower output impedance than the C12 or even the U47.

The first stage uses self bias like the ELA M251 but it has no problem driving the Cathode follower output circuit that has and impedance 20 times lower than the 6072a with fixed bias.


It is a very simple change to make:-

1- disconnect pins 6,7 & 8 from ground. If you heat the socket leg with a soldering iron you can pull them off ground 1 by 1. Worse case is you cut them away and tack a small wire to the remain socket leg.

2-make sure pins 4&5 are tied together and pin 5 is not tied to ground. Also make sure pin 9 in connected to ground.

3- jumper pin 1 to pin 7

4- connect pin 6 to the other side of the plate resistor which is usually a 100K going to pin 1.

5-connect pin 8 to ground through a 270k resistor and tie a 2.2/250v capacitor to pin 8. The other side of this capacitor feeds the red wire of the output transformer.

6-You can remove the plate output capacitor connected to pin 1.


With the CCDA circuit you can easily drive our 6.5:1 iron core dual bobbin transformer fashioned after the BV8 or our BV18 which has a 8:1 ratio and is a slightly larger and lower ratio version of the transformer in the original C12.

Cheers, Dave
 
Hello

Is there any body that could help to make a mod to turm Tbone SCT 700 cardioide into CM47 cadioide
This mod seems appropriate.

I need more information about this mod.

Can someone help me please


the Tbone SCT 700 as a sollder Tube and it seems a little complicated to do the tube connection.

Should i start buy disconnecting the tube?

1- disconnect pins 6,7 & 8 from ground. If you heat the socket leg with a soldering iron you can pull them off ground 1 by 1. Worse case is you cut them away and tack a small wire to the remain socket leg.
2-make sure pins 4&5 are tied together and pin 5 is not tied to ground. Also make sure pin 9 in connected to ground.
3- jumper pin 1 to pin 7
4- connect pin 6 to the other side of the plate resistor which is usually a 100K going to pin 1.
5-connect pin 8 to ground through a 270k resistor and tie a 2.2 ufd metal film capacitor/250v capacitor to pin 8. The other side of this capacitor feeds the red wire of the output transformer.
6-You can remove the plate output capacitor connected to pin 1.
With the CCDA circuit you can easily drive our 6.5:1 iron core dual bobbin transformer fashioned after the BV8 or our BV18 which has a 8:1 ratio and is a slightly larger and lower ratio version of the transformer in the original C12.





CM47
The circuit is designed around the very HiFi 6072 tube and optimized with a two-stage Class "A" CCDA circuit.
This circuit provides a much lower output impedance with more headroom than the original circuit used in the original U47. The lower output impedance provides a faster damping factor when driving the BV8 type transformer, delivering a better transient response than the original design
Selected 6072a tubes are some of the quietest tubes made, and are much less susceptible to microphonics than the metal-clad vf14 tubes used in the original U47 tube microphone.
The dual-bobbin output transformer with its hum bucking configuration and our optimized circuit will deliver > +14dBu with less than 1% distortion.
The output transformer is driven with a 2.2 ufd metal-film capacitor to take advantage of the transformer’s response down below 20Hz
 

Attachments

  • image.jpg
    image.jpg
    111.1 KB · Views: 2
Dave is one of the gems. He’s so humble when you talk with him, yet so experienced and so talented.
If I could have dinner with one of the legends in electronics for audio, it’d be Dave, among a couple other top picks.
Anyway, his mics are stellar. I own the 47 and it’s just perfect on my voice.
He even shares his knowledge generously - he gave me a schematic and told me some component value swaps for a mic mod. Just a star person!
 
And here's the quick Apex460/ Nady1050 mod.

Hi Sevin, upgrading a tcm 1050 to our CM47ve microphone is a fairly simple procedure.
Now, some of the very early 1050 microphones had the 12AX7 tube soldered into the circuit board and replacement of the tube in this version is much more difficult and requires a bit of "extreme" electronic techniques.

It is impossible to heat all 9 pins at once without setting the circuit board on fire. The only way to remove the 12AX7 is to don eye protection, put a rag on the tube and smash it with a hammer over the dust bin. Then cut away all the pieces and unsolder each of the pins one at a time. Then clean out the 9 pin holes with a solder sucker and solder wick. Then you can solder back in a PRE-TESTED 6072a tube. Make sure it is thoroughly tested before soldering it back in!!!!

The later tcm1050 microphones all have a tube socket which makes the tube replacement much easier.

The following changes should be made to the 1050 circuit.

1) C9 & C10 should be removed. They can be cut out and are not required into today's modern world. C9 & C10 where meant as RF filter from the days of Taxi cab's operating their 2-way radios on the upper AM radio frequencies.
These capacitor values were spec'd to low in the 1050. So, C9 & C10 in parallel to ground causes the distortion at 11khz to increase 4 times because of the inherent resonance between these capacitors and the low buget output transformer.
SIMPLY REMOVE THESE CAPACITORS...DO NOT...STRAP THEM WITH A WIRE JUMPER AS THIS WILL SHORT THE OUTPUT OF THE TRANSFORMER TO GROUND.

2) C6 & C7 should be removed. These bypass the cathode resistor in the 1st stage. However, they are not necessary in the 1050 circuit and only necessary in a single stage C12 type circuit.
Once properly modified the 1050 circuit is superior to any single plate output circuit. With C6&C7 removed the circuit is more linear without any LF roll-off. The gain drops 4db but this increases the headroom and signal to noise ratio by 4db.
AGAIN SIMPLY REMOVE THESE CAPACITORS...DO NOT... STRAP THEM WITH WIRE JUMPERS AS YOU WILL SHORT OUT THE CATHODE BIAS RESISTOR.

3) Replace C8 the 1ufd outut electrolytic coupling capacitor with a 2.2ufd metal film. This will place any resonance between C8 and the output transformer down into the sub-sonics where it will not negatively effect the audio signal.

4) Replace the 12AX7 with a 6072a this lowers the gain another 2db but the output impedance on the first plate is much lower. A new Sovtek 6072a selling for around the $20 is more than adequate in this clever CCDA type circuit.

Now, you will have optimized the 1050 circuit which is a 2-stage CCDA (constant current draw amplifier) with an output impedance 10 times lower than the venerable original U47 circuit.

5) Replace the stock economy transformer which has a 10.5:1 turns ratio with our BV8 having a 6.5:1 ratio. The green wire goes to pin 6 and the black wire to pin 5 on the 7 pin connector. The white wire goes to pin 4&7 on the 7 pin.
The red wire is attached to the red wire that goes to the output of C8. Remember, the gain has dropped by nearly 6db in the front end but the BV8 has 5db less loss than the economy 10.5:1 transformer.

6) The economy 32mm (k67) type capsule with its 3-5 micron thick diaphragms is replaced by our AK47 capsule which is a single back-plate w/dual 6 micron diaphragms having identical specs to the original K47 used in the U47 & M49.
The K47 will fit onto the existing saddle but will mount either 2mm closer to the front or the rear. I prefer to move it back 2mm.

7) Next replace Z5 & Z6 sometimes noted D5 & D6 inside the power supply with two 62v Zener diodes. This will drop the power supply B+ voltage to 124v dc + and - 2%. These supplies are usually set for 144v dc and this will place about 70v dc on the capsule. The AK47 and the original K47 capsule seem to give optimum results when they are polarized between 55 and 60v which is what you will get from a 124v dc supply. These are located behind the pattern switch on the board.

That's it!!! If this is all done correctly then the microphone should fire up and sound like a WORLD CLASS MICROPHONE.

I will try and take some pictures of the next 1050 we get in for upgrade.

Cheers, Dave
 
You can remove the tube by unsoldering the pins.
I removed the soldered in tube in two 1050s using a solderpulit and wick.
If you have a good temp controlled iron or better yet a desoldering station it is easy
The stock transformer is ok
There are other things I disagree with in the mod an other things adjustments you can do
 
@kingkorg thanks so much for putting Daves instructions out.
In the past I ordered some stuff from AA for few mods I did and had chance to exchange few emails with Dave.
What a legend Dave is! Never had chance to visit AA forum, it would be awesome to have it back up!


I have 2x SCT-800 coming in for mod and was expecting to be identical circuit as the one from Apex 460 just without pattern circuit but looks I was wrong.

I don't plan to turn it into multi-pattern mic.
So these are the steps Dave recommends

1. Replace stock tube with 12AT7
2. Transformer should be 10:1 ratio
3. Should I Remove C6, C7,C9,C10?
4. Can somebody help with this on (I need to put 200 ufd polystyrene where? C5?)
The stock circuit requires a cathode bypass capacitor so you should fit it with a good 200 ufd electrolytic bypassed with a .01 metal film or poly cap which will improve things a bit.
5. should I change zener diode in PSU to 62V

Thanks
 
More treasure from Dave - Apex 460 mod instructions.

Forget the 251 schematic!!!!!The existing 460 schematic is far superior with just a few simple changes.

There are a lot of myths and misconceptions on Gear Slutz and other Forums.

The plate only circuit is inferior to the 460 circuit when optimized correctly.

The front end (head amp) is identical to the 251 except the cathode bypass capacitors are not required in the 460 circuit.

The ELA M251 circuit has very little if any effect over the sound signature except the very low end a very high end. The sound signature is created by the capsule.

There are a couple of flaws in the original 460 circuit.

1) C9 & C10 should be removed as they cause the distortion to increase at 11khz by 4 times compared to the distortion at 1khz.

2) C6 & C7 can also be removed

3) The tube should be a 60762a

4) The transformer should be a BV11 (5:1) ratio and not the 12:1 T14.

The components in the 460 are much better than components used back in the day. Today we have much quieter and more accurate resistors and capacitors.

I first starting servicing audio tube gear in the 60's and I was always changing bad capacitors. Today we have better power supply regulation and components.


The 460 has a head amp with the same values as 251 but with the more elegant C12 9-pattern selection switch.

The 460 uses the second half of the 6072a as a CF output stage configured as a CCDA (constant current draw amplifier).

I have heard so called Vintage microphone GURU's and Gear Slutz Elitist's tell folks that they have never heard a Cathode Follower circuit they liked.

However, Bill Putnam who designed the LA2 tube compressor drives the output transformer from the cathode. Interestingly, the CCDA circuit in the 460 comes from the output stage of a 1961 State of the Art McIntosh Tube HiFi preamp.

Why does the 2-stage 460 circuit work better than the original circuit?

1) The output of the plate in the first stage is coupled directly into the grid of the second stage and the load on the plate is reduce over 20 times.
Therefore, cathode bypass capacitors C6 and C7 are not required and the gain only drops 4db.

2) The two halves of the 6072 are opposite in current draw so the current being drawn from the power supply remains constant. In the single stage circuit as the current draw changes the voltage on the plate will also change slightly and the gain will vary accordingly. The optimized 460 circuit has much more "punch" and transparency.

3) The output impedance of the cathode is 20 times lower than the plate and a 5:1 transformer can be used instead of the 12:1. There is 8db less loss in the 5:1 than the 12:1.
The result is that decreasing the gain by 4db and decreasing the loss in the transformer increases the headroom and signal to noise.

4) The CF output stage swings the entire B+ plate supply and is not limited by a plate resistor.


The other important thing is to reduce the power supply B+ voltage of the 460 power supply. Change Z5 to a 61v Zener from a 72 volt zener.

This will give you the proper polarization voltage on the CT12.

In the ELA M251 C4 is too small at 20 mfd and should be at least 200ufd to produce a flat response down to 20hz.
However, the German Broadcasting Industry wanted the low end response limited for Radio Broadcasts.

The ELA M251 also used a bypass capacitor (C2) between the plate and ground and this rolls out the very high end 15khz about 3db.

However, in the ELA M251 circuit the amount of roll-off will vary depending on different microphone preamp loads. Today we have much higher impedance preamps.

To duplicate this roll-off with the CT12/460 circuit a 1000pf can be connected from the first plate to ground to give the same curve because the first plate is buffered from the input to the microphone preamp by the CF output stage the HF roll out remains consistent no matter what kind of preamp is used.


However, from a producers point of view I would rather have the HF rise at 14khz and EQ it myself if required.


One of my favourite microphones the Soundeluxe U99 also used this 2-stage CF output circuit.
 
There are a lot of myths and misconceptions on Gear Slutz and other Forums.
And there are some misconceptions here too.
There are a couple of flaws in the original 460 circuit.

1) C9 & C10 should be removed as they cause the distortion to increase at 11khz by 4 times compared to the distortion at 1khz.
The effect of these capacitors is to prevent EMI/RFI ingress. Their value is unduly large, so they should be reduced to 10-22nF, but not dispensed with.
2) C6 & C7 can also be removed
Yes, but the overall gain should be kept into account. I must say the 100nF in parallels with a 100uF is a concession to audiophools, that someone here described as "competition stripes on an Audi".
4) The transformer should be a BV11 (5:1) ratio and not the 12:1 T14.
This is the most debatable point. See later.
The 460 uses the second half of the 6072a as a CF output stage configured as a CCDA (constant current draw amplifier).
CCDA (Constant Current Draw Amplifier) works when the load of the CF is identical to the plate load of the 1st stage.
With a 12:1 xfmr, the load is about 140kohms (with R8=270k and a mic preamp with 2k input Z) an unbalance of 1.4 or 70kohms (with R8=200k and a mic preamp with an impedance of 1500 ohms) an unbalance of 1.4. In order to reap all the benfits of a CCDA, the load must be constant and tuned to the circuit.
With a 5:1 xfmr, the load is about 30kohms, an unbalance of 3.3 (best case)!.
It shows that operating at constant current is possible with a 12:1, if planets are aligned, but utterly impossible with a 5:1.

I have heard so called Vintage microphone GURU's and Gear Slutz Elitist's tell folks that they have never heard a Cathode Follower circuit they liked.
There are also folks who say a cathode bias with decoupling cap can never sound good. Typical audio myth that refuses to die.

2) The two halves of the 6072 are opposite in current draw so the current being drawn from the power supply remains constant. In the single stage circuit as the current draw changes the voltage on the plate will also change slightly and the gain will vary accordingly.
That's the theory, but here it works only if the xfmr has a high enough ratio (min 10:1)
3) The output impedance of the cathode is 20 times lower than the plate and a 5:1 transformer can be used instead of the 12:1.
This is one the myths that must be busted.
Actually the output Z is much lower, but the drive capability is quite low, limited by the high value of the cath resistor, which results in a low idle current. All in all, the headroom is lower by 9dB (R8=270k) or 6dB (R8=200k).
The result is that decreasing the gain by 4db and decreasing the loss in the transformer increases the headroom
No.
and signal to noise.
No. Since headroom is dominated by teh 1st stage noise the S/N is, at best, identical.
4) The CF output stage swings the entire B+ plate supply and is not limited by a plate resistor.
Another misconception. The swing is limited in the negative direction by the cath resistor.
Your affirmation would be true if the load was 100% inductive, same as for a common cath stage.
Today we have much higher impedance preamps.
This is very debatable. There have been various approaches to mic preamp input impedance, in accordance with different countries, different applications...
Many american mic pres had unloaded secondary input xfmrs, which were the recommandation for RCA ribbon mics, resulting in very high input Z at mid and high frequencies.
OTOH, germans, with their predominant use of condenser mics, were perfectly content with a normalized resistive input Z.
 
Last edited:
And there are some misconceptions here too.

The effect of these capacitors is to prevent EMI/RFI ingress. Their value is unduly large, so they should be reduced to 10-22nF, but not dispensed with.

Yes, but the overall gain should be kept into account. I must say the 100nF in parallels with a 100uF is a concession to audiophools, that someone here described as "competition stripes on an Audi".

This is the most debatable point. See later.

CCDA (Constant Current Draw Amplifier) works when the load of the CF is identical to the plate load of the 1st stage.
With a 12:1 xfmr, the load is about 140kohms (with R8=270k and a mic preamp with 2k input Z) an unbalance of 1.4 or 70kohms (with R8=200k and a mic preamp with an impedance of 1500 ohms) an unbalance of 1.4. In order to reap all the benfits of a CCDA, the load must be constant and tuned to the circuit.
With a 5:1 xfmr, the load is about 30kohms, an unbalance of 3.3 (best case)!.
It shows that operating at constant current is possible with a 12:1, if planets are aligned, but utterly impossible with a 5:1.


There are also folks who say a cathode bias with decoupling cap can never sound good. Typical audio myth that refuses to die.


That's the theory, but here it works only if the xfmr has a high enough ratio (min 10:1)

This is one the myths that must be busted.
Actually the output Z is much lower, but the drive capability is quite low, limited by the high value of the cath resistor, which results in a low idle current. All in all, the headroom is lower by 9dB (R8=270k) or 6dB (R8=200k).

No.

No. Since headroom is dominated by teh 1st stage noise the S/N is, at best, identical.

Another misconception. The swing is limited in the negative direction by the cath resistor.
Your affirmation would be true if the load was 100% inductive, same as for a common cath stage.

This is very debatable. There have been various approaches to mic preamp input impedance, in accordance with different countries, different applications...
Many american mic pres had unloaded secondary input xfmrs, which were the recommandation for RCA ribbon mics, resulting in very high input Z at mid and high frequencies.
OTOH, germans, with their predominat use of condenser mics, were perfectly content with a normalized resistive input Z.
Hi,
Thanks for zooming in and sharing details.
My diploma is in Audio Engineering not EL. Engineer and can't comment on circuits flaws but can comment on recording quality.
This formula helped me to build 5 tube mics for under 1000$!!! I find it to be an amazing deal!
All mics sounds really good and can be used in various recording situations with very little processing in mixing stage. I'm sure there is a space for improvement but I'll stop here and focus on making music.
My mic locker is completed :)
 
This formula helped me to build 5 tube mics for under 1000$!!! I find it to be an amazing deal!
All mics sounds really good and can be used in various recording situations with very little processing in mixing stage.
No doubt about it.
It just shows that it is quite easy today to build a decent mic with not a lot of money.
My issue is with people who take advantage of this possibility and claim they have a superior knowledge that allows them to enhance designs, when in fact they have an incomplete knowledge and fill the voids with unsubstantiated cr.p.
I find it's very common within the mic "manufacturers" community.
 
CCDA (Constant Current Draw Amplifier) works when the load of the CF is identical to the plate load of the 1st stage.
With a 12:1 xfmr, the load is about 140kohms (with R8=270k and a mic preamp with 2k input Z) an unbalance of 1.4 or 70kohms (with R8=200k and a mic preamp with an impedance of 1500 ohms) an unbalance of 1.4. In order to reap all the benfits of a CCDA, the load must be constant and tuned to the circuit.
With a 5:1 xfmr, the load is about 30kohms, an unbalance of 3.3 (best case)!.
It shows that operating at constant current is possible with a 12:1, if planets are aligned, but utterly impossible with a 5:1.
Hi abbey,
I’m in the middle of modding 2 GT-2B microphones (same schematic).
Would you suggest keeping the original transformer?
Also wouldn’t driving the 5:1 transformer draw more current from the tube?
Thanks
 
Also wouldn’t driving the 5:1 transformer draw more current from the tube?
Thanks
How and why would it? Or, technically, it is a marginally "harder" load to drive than a 10:1, but easier than a 2:1.

Remember Ohm's law ;)
 
Hi abbey,
I’m in the middle of modding 2 GT-2B microphones (same schematic).
Would you suggest keeping the original transformer?
I would keep the same ratio, or similar. I can't vouch for the quality of the original xfmr.
Also wouldn’t driving the 5:1 transformer draw more current from the tube?
I don't want to give a binary answer, yes or no.
The tube would see a different load, so it would change the current. The cathode resistor is 270 (or 200)k, which would suggest loading it with a similar impedance. A 10-12:1, with a reflected impedance of about 200-250k, is a good fit. A 5:1, with a reflected impedance of 40-50k, is not.
 
Thanks abbey

I've tested the mic with a 6.5:1 transformer (as suggested in the past by Dave from AA) and I'm really not convinced it improves the noise floor! I feel the S/N ratio is the same, but with more gain. Maybe I'm missing something...
 
Back
Top