Head Pre for Denon DL 103

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all
Tiefbassuebertr, in my opinion at your simulation, it is a limited range of R8 values that causes this problem, because the feedback resistor (R4) is 500R, so the feedback is low. In my circuit, R4 is 150R, so the feedback is higher, thus, circuit is more prone to instability.
Or if you have to advise me for some other sort of cure.:confused:
Regards George
Joachim Gerhard: i can send you some anti oscilation coils that you inset in series with the cardridje


Please note: R4 changed not only the AC conditions like wished gain factor and high frequency behaivor !!! Through large changed DC conditions if R4 = 150R, I don't get simulation result, only unspecific error message.
Only if I adjust various other values (especially R2) I get again simulation result, but to get perfect roll of without peaks (i. e. results without oscillation) further action must be taken.

Please note additional, that the simulation don't respect any parasitic effects through PCB layout, the kind of wire and the quality of power supply (i. e. resistors, capacitors and power supplies have still ideal character).

If you are beginner, it is very hard to find satisfy solution without any unwanted effects (e. g. long distances of the base line wire from input load resistance to LM394 are very critical).

Joachim Gerhard:
I have heard about anti oscillation coils for microphone head amps. But I have never heard about this one for RIAA head amps to prevent parasitic oscillation, except if I have unwanted reception of AM short wave broadcast stations - go to
http://www.diyaudio.com/forums/anal...unwanted-reception-broadcast-sw-stations.html
Unwanted radio station reception and parasitic oscillation are not the same thing and thus not basicly similarities there are present.
 
Last edited:
Member
Joined 2002
Paid Member
Hi all

This was another day devoted to the head amplifier.

Dear friends, finally I got rid of the oscillation by connecting a 1nF capacitor across the base and collector of LM 394. Now the HF –3dB point is at 700kHz.
I think this is a technically better solution compared to the capacitor across the feedback resistor (which cap I have now removed). The cap at the feedback path was suppressing an existing oscillation, while the cap at the LM 394 prohibits the oscillation.

Actually placing a cap across b-c of the first transistor is an old trick to kill RF interference, but I saw it suggested here as well http://www.diyaudio.com/forums/solid-state/25172-baxandall-super-pair.html#post293664

The head amp is now dead quiet. I could not measure anything else than 0.1 mVpp HF noise at the output with input shorted.
 

Attachments

  • Final Schematic with changes.JPG
    Final Schematic with changes.JPG
    71.8 KB · Views: 561
  • Final Freq Resp 5-12-09.JPG
    Final Freq Resp 5-12-09.JPG
    156.2 KB · Views: 549
Member
Joined 2002
Paid Member
Joachim
I appreciate your offer. I solved the problem with the capacitor, which I do not see causing any side effects. Thank you very much.

Tiefbassuebertr
Unfortunately I am not very familiar with simulations.
If I can help you by measuring DC voltages on the circuit, I will do it gladly.
Yes, you are right that the parasitics can not be estimated in advance, that’s why I opted for this kind of construction, so as to minimize trace lengths. But it is a difficult and very time consuming construction method, and it doesn’t adapt itself to component changes.

Right now the Head amp plays music in a very satisfying way. I do not have another head amp to compare it with, but I like what I hear.

I will use it to make comparisons between an unmodified and a modified DL 103, for which I will inform you. Then, I will house it in a screened enclosure, and then build another head amp (Hiraga?, SL-10 ?) just to see for any differences.

Thank you all for your help.

I'll be back tomorrow

Best Regards
George
 

Attachments

  • Final close-up.JPG
    Final close-up.JPG
    75.2 KB · Views: 523
  • Final Temp set-up.JPG
    Final Temp set-up.JPG
    131.8 KB · Views: 506
Joachim
I appreciate your offer. I solved the problem with the capacitor, which I do not see causing any side effects. Thank you very much.

Tiefbassuebertr
Unfortunately I am not very familiar with simulations.
If I can help you by measuring DC voltages on the circuit, I will do it gladly.
Yes, you are right that the parasitics can not be estimated in advance, that’s why I opted for this kind of construction, so as to minimize trace lengths. But it is a difficult and very time consuming construction method, and it doesn’t adapt itself to component changes.

Right now the Head amp plays music in a very satisfying way. I do not have another head amp to compare it with, but I like what I hear.

I will use it to make comparisons between an unmodified and a modified DL 103, for which I will inform you. Then, I will house it in a screened enclosure, and then build another head amp (Hiraga?, SL-10 ?) just to see for any differences.

Thank you all for your help.

I'll be back tomorrow

Best Regards
George

Congratulation!
Follow residual unwanted effects you can easy remove:
1) the peak between 2 and 10 Hz (enhance capacity value of C1 to 68 uF + bypaß MKP 1uF parallel for better behaivour at higher frequencies)
2) If you use your head amp together with screened enclosure, perhaps you can remove your Cdom of 1nF between base and collector of LM394 without oscillation effects (lower risk of TIM). This Cdom reduce the open loop bandwidth massive.

It's not a problem that you are not familiar with the p-spice simulation. But you now can compare both DC voltage values from simulation of me (you can find it inside of my schematic) and that one at your diy head amp device. This makes troubleshooting more easy and ensures at least a similarity of the voltages in theoretic and real world. If there are geat differences after adjust of R2, there must be an error.

I am also not a big fan of simulations, because I don't perform the verify of the P-Spice parameter and therefore I must hope to serious information of the transistor manufacturer. Additional some devices would be taken as ideal (resistors, capacitors, voltage sources). In real life this is never the case.
Nevertheless, I must say, that tool it is a great resource for me whether there are fundamental design errors and the risk of unwanted effects or not.

P.S. You haven't forget the 10 ohm resistor in series to the capacitor C2 (200 uf) in your head amp ?? Please let me also more about your power supply, that is in use.
 
Last edited:
Member
Joined 2002
Paid Member
Hi all
As promised, here are DC and AC measurements of the circuit.
AC measurement refers to 1KHz input signal (I forgot to mention it on the drawing, sorry)

tiefbassuebertr
Thank you for the suggestions.

P.S. You haven't forget the 10 ohm resistor in series to the capacitor C2 (200 uf) in your head amp ??

There isn't such a resistor in the drawing. What do you mean?

Best Regards
George
 

Attachments

  • AC Data.JPG
    AC Data.JPG
    98.2 KB · Views: 198
  • DC Data.JPG
    DC Data.JPG
    90.3 KB · Views: 496
Last edited:
tiefbassuebertr
Thank you for the suggestions.
There isn't such a resistor in the drawing. What do you mean? Best Regards
George
Right. But in much other sziklai schematic drawings, even by National Semiconductor applications this resistor is present.
The aim of this serial resistance is to reduce the open loop gain and enhance the open loop bandwidth. The risc of oscillation is lower and hfe spreads of the sziklai transistor (in your schematic the BC416C) contribute less critical.

An second thing I have see: you create your input cap as a unipolar (bipolar) version through two antiserial connected electrolytics. Why?

On base of LM 394 there is a positive voltage. At the hot end of your moving coil of DL103 (respective hot site of input load resistance) 0 volts. So I would clearly say, the pos. pole of capacitor must connected to the base of LM394.
 
Member
Joined 2002
Paid Member
Hi tiefbassuebertr

I am glad that you provide me your knowledge.

I had to go and read about the sziklai pair (first time that I heard this name):eek:.

After this reading, I understand that the resistance you mention is to be connected between base and emmiter of the PNP "slave" transistor.
Is this correct?

An second thing I have see: you create your input cap as a unipolar (bipolar) version through two antiserial connected electrolytics. Why?

Rumour has it that they behave better in audio circuits than single polarized. So I followed.
Is this another myth?:confused:

On base of LM 394 there is a positive voltage. At the hot end of your moving coil of DL103 (respective hot site of input load resistance) 0 volts. So I would clearly say, the pos. pole of capacitor must connected to the base of LM394.

According to the rumour above, such a unipolar capacitor version can have as outer electrodes either the 2 negatives or the 2 positives with no difference for the connectivity to the external circuits.
Yet another myth? :headshot:

Thanks in advance (for the coming response):)

Regards
George
 
Hi all

I build a lot of different MC PrePre`s in the past.And swapped all low noise pnp`s in my hiraga prepre.

Best results with:

2SA1038 from Rohm sounds much better than the 2SB737 / BC560/ MAT Type
The biggest improvement was a pair of super e black gate nx from base to ground.

For resistors I use shinkho tantalum for couplers mundord silver

I build 5 units for my buddys and the TX-103 Billington stepup and Audio Note silver stepup (the big one for $10000) are sold on ebay

I build the pcb flat on spruce wood and it sounds much better than every preamp or stepup transformer I used in the past

regards Tony
 

Attachments

  • DSC_0258[1]_1.jpg
    DSC_0258[1]_1.jpg
    101.1 KB · Views: 268
Hi tiefbassuebertr

I am glad that you provide me your knowledge.

I had to go and read about the sziklai pair (first time that I heard this name):eek:.

After this reading, I understand that the resistance you mention is to be connected between base and emmiter of the PNP "slave" transistor.
Is this correct?



Rumour has it that they behave better in audio circuits than single polarized. So I followed.
Is this another myth?:confused:
According to the rumour above, such a unipolar capacitor version can have as outer electrodes either the 2 negatives or the 2 positives with no difference for the connectivity to the external circuits.
Yet another myth? :headshot:

Thanks in advance (for the coming response):)

Regards
George

The resistor connected between base and emmiter of the PNP "slave" transistor dermines the C-E idle current through the master transistor (in your case it would be the LM394, but please note, in your case you have different pads for AC and DC !!! that means, the idle current would be at whole determine R5/R6 and R3)
The resistor in series to the elko C2 (200 uf) is actually the serieal resistance of the emitter of slave transistor, when there are only one pad for DC and AC.
It reduce the HFe of the slave and the open loop gain of the whole preamp.
Please use follow keywords by google:
"Sziklai Darlington" and "Current Feedback Pair" respective "CFP"
here some examples
Electronic Design Welcome
http://www.ampslab.com/PDF/sziklai_cv.pdf
http://www.tunerlux.com/files/Insida_f^rdig.pdf
Electronic Design Welcome
http://www.diyaudio.com/forums/solid-state/8935-yet-another-amp-design.html
http://www.diyaudio.com/forums/soli...-diamond-buffer-sddb-power-output-stages.html


Rumour has it that they behave better in audio circuits than single polarized ??? I have never heard about this. Please send me URLs with this claim. But I have heard, that this could be a solution in case uf unipolar situations (i. e. 0V at both ends). In such cases I recommend either foil caps (size is problem sometimes) or high voltage Elkos (>100V) even if the max. voltage is low. Some developers put on a diode parallel to limit the max voltage to 0,7V.
For me the best way is a polar situation or the use of foil capacitor (MKP, MKT, MKC).
 
nice work Tony

I would be interested in some more pics.
The PrePre on the pics is made from Mahogany wood I think ( nix Fichtenholz). The inner tin, is it brass or copper? Very nice brass front.
What info are the two windows showing?
I have used 2SB737 Rohm and 2SA1316 and 2SC3329 before.
What seize have the base to ground caps? Can you describe the improvment of sound more detailed?

My actual prepre sound somtimes harsh with the "SS" at the end of the words. I am using Multicap RTX 1,5uF and 0,1 Russian Teflon cap in parallel.
I really don't know if it is the PrePre or the rest of my chain ( Kaneda RIAA and LINE + 8WAtt Hiraga + FE103 BR, all amps with Shinko.

Thanks for your answer, Tony

Regards

Sam
 
Hello Sam

The Mahagony painted block of wood (inside) is spruce wood
The inner thin is copper.
The caps from base to ground are Black Gate Nx 100uF 6.3V in antiparallel configuaration (2 per channel)
The sound is more powerfull and with better resolution and natural colours
I forgot a 100 ohm resistor between the 2 bases and the Black Gates are connected to the base of the signal transistor.(The Nlack Gates are not on the pics)
And I forgot the most important thing 2 pc of optima redtop batteries, sounds much better than Sonnenschein Dryfit or Panasonic.
For coupling is Mundorf silver or Vcap the best,but Vcap is too expensive for this aplication
The Instruments are for adjust the dc offset on the input, the 9V battery inside are the supply for the instruments.
I can give you my telephone number when you are interrested of further informations.

regards Tony
 

Attachments

  • DSC_0260[1]_1.jpg
    DSC_0260[1]_1.jpg
    111.3 KB · Views: 426
  • DSC_0264[1]_1.jpg
    DSC_0264[1]_1.jpg
    84 KB · Views: 413
Member
Joined 2002
Paid Member
Hi tiefbassuebertr

Ref your post #50
Thank you again for the links.
A lot to studying and thinking. I'll come back to this some days later.

About the bipolar caps, a lot of URLs, but basically they all refer to Cyril Bateman articles in Wireless World magazine.
A sum-up is here: http://www.diycore.com/patrick/files/electronics/capacitor/Cap_fallacies.pdf

Hi pdul
Thank you for contributing.
Have you built this?
With this PSU?

Best regards
George
 
@ Tony: phone number would be nice.

Vielen Dank für dein Angebot Tony. Ich würde gerne mit dir sprechen, und auch mir dein Werk mal anschauen. Ich wohne nur 70 km südlich in Greding an der A9.

Bin die nächsten2 Tage unterwegs würde mich dann melden. Nummer bitte an mich bitte per Email.

Grüße

SAm
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.