I'm definitely not that, I know just enough with LT to get by 🙂 If you want to ask about Germanium models or to ask for help in creating a model in the LT thread then that is fine. If you actually have the transistors available then a prototype (or breadboard lashup) might be useful to get an idea how it might work in practice.
I'm very interested in building this schematic, it seems to be the most interesting in this field.
Only the NPN transistor is missing in my inventory by now....I don't have also the ACY21, but I think OC44K will work fine instead.
Even if I have a bunch of good AC125 (Tungsram), I wonder if the CCS won't be the same with an 7805 and 100R resistor, the 7805 wired as CCS. Could this affect the sound? In my opinion, it should not, eventually I will place in parallel some capacitor - let's say 47uF - to have a supplementary feedback. I used such solution in tube amps, as cathode resistor for EL84 and it worked like a charm. Well, there was LM317 and some 20-ish ohms resistor, but the idea is the same. Also there I used some 100uF capacitor in parallel with the LM317.
So, instead of R14, Q4, Q5, and R10 - I should use am 7805 - or even 7905 - with an 100R resistor, as the datasheet of 7805 says - for a 50mA ccs.
Any opinions?
Only the NPN transistor is missing in my inventory by now....I don't have also the ACY21, but I think OC44K will work fine instead.
Even if I have a bunch of good AC125 (Tungsram), I wonder if the CCS won't be the same with an 7805 and 100R resistor, the 7805 wired as CCS. Could this affect the sound? In my opinion, it should not, eventually I will place in parallel some capacitor - let's say 47uF - to have a supplementary feedback. I used such solution in tube amps, as cathode resistor for EL84 and it worked like a charm. Well, there was LM317 and some 20-ish ohms resistor, but the idea is the same. Also there I used some 100uF capacitor in parallel with the LM317.
So, instead of R14, Q4, Q5, and R10 - I should use am 7805 - or even 7905 - with an 100R resistor, as the datasheet of 7805 says - for a 50mA ccs.
Any opinions?
The OC44 should work. Given the low supply voltage and (relatively) high current of say 6ma in that first stage you may need to tweak the biasing to get symmetrical clipping,
The CCS using germanium works with a low voltage overhead, I don't think the 78/79 regs would work tbh.
This should click and run if you use LTspice.
The CCS using germanium works with a low voltage overhead, I don't think the 78/79 regs would work tbh.
This should click and run if you use LTspice.
OK, I give up the idea of 78xx CCS, I wasn't aware of the low supply voltage, but you must be right....maybe a LDO regulator could solve it, but this complicates unnecesarily the schematic. Let it be 4 more germaniums then.
What should we have in sight while replacing that first transistor? I thought about OC44 being known as a low noise one, should I use some "driver" or more collector-current capable transistor? Now I see the case format of ACY21, it seems rather medium power than input stage transistor...the only ones in this case I have are two Polish transistors TG50....salvaged from an old Unitra portable radio. As I remember, they were in the final stage there.
I don't know how to use LTSpice....
What should we have in sight while replacing that first transistor? I thought about OC44 being known as a low noise one, should I use some "driver" or more collector-current capable transistor? Now I see the case format of ACY21, it seems rather medium power than input stage transistor...the only ones in this case I have are two Polish transistors TG50....salvaged from an old Unitra portable radio. As I remember, they were in the final stage there.
I don't know how to use LTSpice....
Have a look here. The attached file should click and run and that gets you a long long way into figuring LTspice out.
https://www.diyaudio.com/community/...rom-beginner-to-advanced.260627/#post-4025305
And here:
https://www.diyaudio.com/community/...-2023-version-of-ltspice.394676/#post-7240021
I ending up having to use what I had rather than what I might have wanted to use. So its very much a case of try what have available I think. As you say, noise is a consideration and if you look at the first first few posts in the thread you will see I tried an AF116 for the front end and while it had benefits in speed it was audibly 'hissy'. So try what you have 🙂
More fun using Ge for CCS also (and needs less critical transistors). I also vote against LDO since LDO current drain is no-linear with low voltage levels over the internal IC pass trasistor.
Yes, that I was about to ask: if I could use the lower beta ones...I have 10 AC125 but they are not equal, some have decent beta and a fiew are under 30...I will re-measure them and sort accordingly...and if my judgement is right, I can use a lower Ic transistor in the position of Q5...in that case that's all set for me....Maybe for the NPN I will use 106NU70, old Czech transistors made by Tesla or two Romanian made EFT373. Sort of AC176.
In the first position I will put sockets and roll more versions, OC44K, also some RF AF1xx or other options till I like the sound.
In the first position I will put sockets and roll more versions, OC44K, also some RF AF1xx or other options till I like the sound.
A first attempt at PCB for the schematic in the first post. For colleague @carageae
I'll work on it a little more for optimization and I'll post a final version.
This is done in a hurry.... 😎😎😎😎
I'll work on it a little more for optimization and I'll post a final version.
This is done in a hurry.... 😎😎😎😎
Last edited:
Sounds like a good idea. Q5 can be the lower beta one, Q4 the next higher (save the higher for Q3), and Q6 not need to have a lot high beta.
Thank you, @Julian RO !
In the next days I will measure and select the AC125's and I will try to make a hand-drawn PCB. If succesfull, maybe we'll make some industrial PCB's for this.
Before using the lower hFE AC125 transistors I think I cand tray some small signal germaniums, which I don't intend to use anymore, in that position I prefer the silicon ones. If they fail, I will replace with the weaker AC125...but even in the 2-nd quality grade they are still good ones, being NOS Tungsrams...
In the next days I will measure and select the AC125's and I will try to make a hand-drawn PCB. If succesfull, maybe we'll make some industrial PCB's for this.
Before using the lower hFE AC125 transistors I think I cand tray some small signal germaniums, which I don't intend to use anymore, in that position I prefer the silicon ones. If they fail, I will replace with the weaker AC125...but even in the 2-nd quality grade they are still good ones, being NOS Tungsrams...
Turn - on voltage of 0.3 Volt for germanium, 0.7 Volt for Silicon, if my memory serves me well. So overcoming minority carriers is that much lower (0.4 volt lower). But that as Mooly says, seems so long ago. I believe Germanium is inherently lower in noise too, assuming it is temperature controlled - used in the front end of domestic satellite dishes? Or have I, like Germanium been, supposed by modern developments.
Those voltages sound about right (0.6v was always the figure when I started out but modern processes and doping have seen that rise a little... one reason why substituting old driver and output transistors sometimes puts the bias adjustment out of range) and I think a lot of germanium will start to come into conduction quite a bit below 0.3 volts.
.... Gallium Arsenide?
0.2V and 0.7V were the values I remember from long ago. Silicon transistors in practice are often in the range 0.6 to 0.7V for small signal use. There is no defined "turn on" voltage with BJTs, the current is an exponential function of Vbe until resistance starts to dominate. Ge devices have a tendancy to undergo thermal run-away if not carefully used (and can be destroyed by soldering temperatures, unlike silicon devices in my experience). For Ge the leakage currents get large once the device has warmed up a bit, making Si much easier to design with.
I also asked about this transistor, it should be chosen experimentally, it's in fact nothing special, but the author stated that other "better" transistors performed noisy...i will place some pins from a broken DIP8 IC socket, try different transistors until I like the sound, then solder the final candidates in place. Germanium transistors have thin terminals and will fit tightly on those pins, I did that before...use sockets like this image, not the "spring tape" ones...jus brake the plastic and you have 8 useful test pins. Other solutions exist, of course, I just show here what I tried and worked before.
If somebody wants to stick to the originals, even if @Mooly said it was a random pick, you can find those transistors here:
https://www.langrex.co.uk/products/acy21cv7439-newmarket-germanium-transistor/
If somebody wants to stick to the originals, even if @Mooly said it was a random pick, you can find those transistors here:
https://www.langrex.co.uk/products/acy21cv7439-newmarket-germanium-transistor/
