EMI TG mic preamp schematic

[ruffrecords]

This schematic seems to be compatible with the number of components in the TG 500 module.
It remains only to know the right value of resistors.
For the gain we can use a pot intead of R4?

R4 would be the fine gain pot, arranged to vary the gain from 20dB to 30dB as in the original.

Cheers

Ian

 

[Strawtles]

Hi Ian, but what's the right values for the resistors?

 

[Liutmod]

Yes Ian, can you tell us what are the values for the resistors and caps in the circuit you have posted?
Thank you 

 

[ruffrecords]

Yes Ian, can you tell us what are the values for the resistors and caps in the circuit you have posted?
Thank you 

That is the $64K question. There is a whole range of values they could be depending on the load the amp will be driving and the source feeding it. There is no single answer so I will have to make some sensible guiesses.

Start with R2 and R3 the same value. That will give 6dB gain with no gain pot and we can increase it easily to 20 or 30dB.  We want the collector of Q2 to be half the supply voltage for maximum output swing, so if the supply is 24V then Q2 collector needs to be 12V. The values of R2 and R3 can now b determined if only we knew what collector current we need in Q2. We don't need to drive a heavy load so 6mA will probably be enough. This means R2 +R3 = 2K (12V/2K = 6mA) so R2 and R3 are both 1K.

Q1 emitter is now at 6V so its base needs to be 6.65 volts or thereabouts. R5 and R6 for a pot divider across the supply to provide this 6.65V but they need to be large enough not to load the input. If we allow 0.1mA in this chain then the total value is 240K with a 24V supply. So R6 about 66K and R5 is about 175K - we could probably use a 68K and a 180K. The load these resistors present to the input is just under 50K - probably big enough.

Why did we choose 0.1mA through the bias resistor chain??  - because we are going to choose 0.1mA collector current for the first transistor because this will give lower (but probably not the lowest) noise and a good rule of thumb for bias stability is to make the bias chanin current equal to the collector current as long as the transistor hfe is more than 100 (which it is for a BC109). We know the collector of Q1 is only 0.65V below the rail because it is directly connected to Q2's base. For 100uA in Q1 we therefore need to set R1 to 6.65K. ^.8K will probably be fine - the collector current will just be a bit different and the dc negative feedback will sort out the dc conditions.

We have ignored the fact that Q1 collector current flows through R3 thus upsetting our assumptions about bias conditions but it is only 1/60th of Q2 collector current so the effect will be small.

Thies values should drive a 10K load with not problem. For a 3dB point at 3Hz with a 10K load, C3 needs to be at least 5uF. We can use 47Uf andmove the 3dB point down to 0.3Hz and forget about bass phase response.

At the input, the bias network looks like 50K but this is also in parallel with the input impedance of Q1 which is likely to be of a similar value (too complex to explain). So assume input looks like 25K. Using the same reasoning as for C3, this means we should start with 22uF for C2.

For the gain resistor R4 we know we want a maximum of 30dB of gain which is 32 times so R4 in parallel with R3 needs to be about 1/30 of R2 which is 33 ohms. For 3dB drop at 3Hz,  C1 therefore needs to be  1600uF which is a tad on the big side. So we will have to reduce the current in Q2 to 3mA and increase R2 and R3 to 2K each. R2 increase to 66 ohms and C1 will be OK at 1000uF.

As you can see, this design is not without its compromises simply because we are trying to make a decent mic pre out of just two transistors. There are some simple modifiactions you can make to this circuit to make it work better but it is worth trying the above values in a simulator and seeing what you get.

Cheers

Ian

 

[Fablab]

Considering how  different values of the components may vary the results, I think the best thing would be , if someone who has this preamp can trace the circuit of the first stage (at least), as suggested by Ian some posts back.

 

[Strawtles]

Hi Ian, I put the values in the schematic you posted but I don't understand what of the two transistors is the BC184 and what is the BC212
Where I can put the 12 positions gain swich?
Thank you

 

[ruffrecords]

Q1 is the BC109.

The gain switch in the original is actually an attenuator between the transformer and the input of the preamp.

The fine gain control replaces  the gain resistor.

Cheers

Ian

 

[Liutmod]

Ian, if possible I have two question for you.
The first is about the power supply for this circuit. It is single or dual ?
The second is about the value for the 12 resistors in the gain rotary switch
Thank you

 

[ruffrecords]

Ian, if possible I have two question for you.
The first is about the power supply for this circuit. It is single or dual ?
The second is about the value for the 12 resistors in the gain rotary switch
Thank you

The sketch schematic I drew uses a single rail but I believe the original uses a dual rail.

The original had a 1:3.16 transformer with secondary taps at -5dB and -10dB and then a stepped pot from the -10dB tapping for higher level inputs up to +20dBV. I do not know if the Chandler uses the same technique, the same gain range or even the same step size. According to the Sound On Sound review, the coarse gain switch only has seven positions from 50dB gain down to 20dB. This probably avoids having to use an input transformer with a tapped secondary. The two transistor amplifier has a nominal gain of 25dB and the transformer has 10dB which is a total of 35dB. So a -50dBu input will leave the preamp at -15dBu. A -20dBu input will need a 30dB attenuation for the same output level so this sets the maximum loss in the coarse gain switch.

If we want the mic input to look like 1.5K then a  1:3.16 input transformer needs a secondary load of 15K so this is the maximum value of our attenuator (ignoring the input impedance of the preamp itself). So we need a seven step attenuator of 5dB steps and a total resistance of 15K. A quick visit to a stepped pot calculator web site gives the following resistor values:

6K8, 3K9, 2K2, 1K2, 680,390 and a bottom resistor to 0V of 470 ohms.

Cheers

ian

 

[ruffrecords]

Here is a somewhat improved version of the two transistor preamp. Improvements include:

1. PNP input transistor for lower noise
2. Higher input impedance
3. Higher current in second transistor for lower distortion
4. Higher gain setting values for more sensible associated capacitor value

The transistors just happen to be ones available in LTspice with the right sort of characteristics.

Cheers

Ian

 

[Strawtles]

Thank you Ian 

 

[dukasound]

My 50 cents about this dissscussion. 
Here are two different concepts  TG12345 and TG12428.
TG12345 with input transformer 1:3,16 ratio (with secondary load 8,33K and special winding like Sowter made), three BC109 transistors  in first stage with fine trim +/-5db (20-30db gain) and second stage with BC109 and ACY21 (germanium) trimed do give 10db with output transformer who is 1:1,78. Overal gain can be 50db, and work on +/-20V.

TG12428 with 1:2 input (Neve style with about 6db gain and probably 5K secondary load with gain switch), one or two stages based on BC109 and BCY71 (BC184 and BC214) with additional 40db, than output stage who is used almost in all TG circuits with BC109, BCY71 and bigger 2N4910 and 2N4898 who gave 20db and output transformer  50/600ohm, ratio 1:3,46 who gave another 10db. Overal gain is about 75db.  Too this kind circuit must have some voltage dropper  viewed on limiters and EQs who 28V drop to 24V or  20V depends of design. This is based on Candlers type TG-2 and observation of all EMI TG stuff. I am pretty sure about all expect first or second stages. Too, I saw Variable Gain Amp in TG12411 and I am pretty sure about same or similar design on new TG-2.  All parts are same exept 200u (on old schematic) instead 22uF on new design. 

 

[dukasound]

Variable Gain Amp in TG12411 ???
There are another two transistor BC184 and BC214 but are they work both for preamp stage or one for preamp and another for  voltage dropper I dont know.

 

[RuudNL]

What could be the idea behind the transistor in the 'voltage dropper' section of GroupDIY TG2.png?
I could imagine that it is a sort of 'slow starter' for the +20 V supply voltage?
(After power on, C126 will start to charge, causing VT109 to conduct.)

 

[Strawtles]

Thank you

 

[ruffrecords]

Looks like my original idea for the variable gain amplifier was surprisingly close to the mark. I will have to model their circuit to see how it performs.

Cheers

Ian

 

[Liutmod]

Hi Ian, any new about your test with the circuit?

 

[ruffrecords]

Hi Ian, any new about your test with the circuit?

I just tried it on PartSim the online simulator. Frequency response looks OK at 30dB gain but I had to increase the gain capacitor to 470uF to get a decent bass response. I could not get PartSim to accept spice models for the BCY71 and BC109 so I used ones they hade that were reasonably similar. I have not tried a distortion measurement yet as it is not obvious how to do this on PartSim.

Aside: I love LTspice. It is free, runs spice models from just about anywhere and is easy to work with. However, you need a PC to run it on and I don't always have one to hand. So an online would be great. PartSim is OK but a bit counter intuitive and rather picky about spice models. Anyone know a better on line sim?

Cheers

Ian

 

[Strawtles]

Hi Ian, about the gain rotary switch you posted: 6K8, 3K9, 2K2, 1K2, 680,390 and a bottom resistor to 0V of 470 ohms.
Please, can you help me to understand how to connect the 470 Ohm resistor?

 

[ruffrecords]

Hi Ian, about the gain rotary switch you posted: 6K8, 3K9, 2K2, 1K2, 680,390 and a bottom resistor to 0V of 470 ohms.
Please, can you help me to understand how to connect the 470 Ohm resistor?

Attached is a quick sketch of what I mean. Apologies for the scruffy hand drawn schematic but I tried doing it on a CAD program that only has a 6 way switch and I could not fit the resistors in so I had to keep moving them about and eventually gave up and used pen and aper.

Cheers

Ian