Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 10:54:59 AM

Login with username, password and session length

Recent Posts

 

National GA920P Schematic

Started by galaxiex, September 13, 2015, 12:33:41 PM

Previous topic - Next topic

galaxiex

#15
Quote from: J M Fahey on November 11, 2015, 06:56:38 AM
QuoteI guess what also confused me is the cap, C7, where one would expect to see a load resistor.
I realize now that the tank transducer coil is the actual load, and the cap is simply across the coil for....
Stability?
Frequency compensation?
Other?
Correct.
The collector load is the drive coil.
Collector current passes straight through that coil, and is modulated by signal driving the transistor.
Since that current is "pushing" the drive magnet towards one side by a fixed amount, depending on coil turns, actual current, magnet strength, distance, etc, varying (modulating) said current will make the magnet mofe further or closer, vibrating/twisting following audio input.
Said twisting runs along the spring, takes some time to reach the other end (delay time, usually between 30ms to 80 ms, depending on tank) and produces the reverb effect.
Current  (at idle) is between 8 and 10 mA , a lot considering the coil has hundreds of turns.
QuoteIt just looks odd to see they drive the tank coil directly without a transformer.
What for?
Maybe you mean in *tube*  circuits, where they must adapt tank low input impedance (in old Fenders around 8 ohms) to high tube impedance (12AT7) , but not here.
In fact, most modern Tube amps also drive the tank coil without transformers , either straight with an Op Amp or with a more powerful discrete amp .
Our friend Phatt hits them with a sledgehammer (TO220 transistors fed +/- 35V rails  :trouble  )
Quote
I'm so *used to* seeing a transformer at the input of the tank.
Browse a couple dozen modern schematics.
Modern meaning anything from a late 60's Kustom on  :lmao:
QuoteAlso the tank transducer coils being almost equal resistance, I had not come across before.
Same.
In a Fender Black/Silverface reverb reverb drive is a small 1W power amp, parallel single ended 12AT7 , and the tiny output transformer can drive a small speaker, so it was plate to speaker wound, so input impedance was around (nominal) 8 ohms, typical speaker values, while output transduces uses many turns of very fine wire to have higher output voltage, so high impedance.

Now an Op Amp or a tiny transistor are not happy driving an 8 ohms load, but feel fine in the range of, say, 200 ohms to 2k to 10k .
Now, once the reverb factory set up the automatic winder for high impedance (a couple kohms) coils and it's churning, say, 1000 coils an hour, it's a very attractive idea (for the production manager)  to be able to use the exact same coil at the driving end.

In a nutshell:
* output coils will be as high impedance as possible to increase signal.
Not actually very high because wire becomes impossibly thin and breaks all the time (it sure ruins the manager's day) but a couple kiloohms is an acceptable compromise.
* input coils will be around 8 ohms for Fender type tube+transformer drives
* around 2000 ohms for single Op Amp driven tanks .
Regular Op Amps (741/RC4558/TL072)can't drive (well)  loads below 2k .
* around 150/200 ohms for small transistor boosted Op Amps (check the Lab Series reverbs).


Here they use a full sized power amp, fed +/-25V, with TIP29/30 output transistors (1 Ampere instead of typical 5mA available from an Op Amp) , so loud that they use an 8 ohms input tank but they must attenuate it with a series 56 ohms resistor.

Phatt does a similar thing but with +/- 35V rails  :loco  and mid/high impedance tanks.

Otherwise the tiny drive coil would burst in flames ... not kidding

QuoteWhen I first drew that circuit my first thought was...
WTH! this can't be right!....
Does it look better now?  ;)

As of the parallel cap, the coil is an inductor, and in theory it will have infinite impedance at some high frequencies ... the transistor will not be very happy ... the parallel cap avoids that.

Wow JM! Thanks much for all that!  :)

I freely admit to having little understanding of the design of circuits.
Some (simple) stuff I can figure out or intuit from surrounding circuitry.
(I guess it's all simple when you have years of experience/training...)
I'm mostly a tinkerer, tho I do study and try to understand this stuff.  :)

I very much appreciate you explaining and putting in language anyone can understand how this works and the reasoning behind it all.

This is what the "textbooks" lack, and they often assume a high(er) degree of knowledge/training than what I possess,
so it can be difficult for me to get "understanding" from textbooks.

Not to mention, the texts are often rather thick with math.
I can use Ohm's Law and often do, and simple stuff like RC time constant.
But when I want to know how to bias a NPN bipolar transistor and the textbook goes into pages and pages of esoteric (to me) calculations and formula I get cross-eyed and dizzy.

I just want a (hopefully) simple, "plain language" explanation of what the parts in the circuit do,
so I can *by-guess and by-gosh* make something that works.
The texts are for a engineering point of view... I'm just tinkering... I end up copying others work and tinker from there.

Yeah, older Fender tube reverb circuits are what I've mostly looked at.
I will look/study/pay attention to, some "modern" circuits.  ;)

What you say about the tank coils certainly makes sense.
I get the impedance thing, and also how the "suits" like to make everything easy and cheap to manufacture.

Ya, it all looks better now.... Amazing how just a little understanding can clear up so much!

Thanks again!  :)
If it ain't broke I'll fix it until it is.

g1


galaxiex

Thanks g1! That is indeed interesting.

I notice that document shows a Blues Junior partial reverb schem which does not use a matching tx.
The tank is directly driven by a TL072.
Interesting... because I have a Blues Jr. and never noticed that before, or,
more like, I wasn't paying much attention to the reverb circuit whenever I looked at the schematic.  ::)
If it ain't broke I'll fix it until it is.

galaxiex

Necro bump for a small update.

I created a crude board layout for this amp.
If it ain't broke I'll fix it until it is.

galaxiex

Also slightly updated the schematic.
If it ain't broke I'll fix it until it is.

clintrubber

Thanks for sharing this info on the GA920P, much appreciated!

Have one here as well, no idea if it still works, I've never plugged it in after taking it home from a fleamarket (never dared to..., have to check it for safety first)

Mine didn't come with a schematic.

Was wondering, any info on the reverb tank they used ?  Input- & output-impedances ?
Then possibly another circuit could be used to drive the tank.

I'll admit I was intending to use the enclosure forsomething else, but who knows it's a fun amp by itself.

Bye/thanks   

clintrubber

Quote from: clintrubber on November 01, 2017, 09:05:23 AM
Was wondering, any info on the reverb tank they used ? Input- & output-impedances ?

Oops, apologies, just saw the 225R & 222R indications in the last version of the schematic. Are these as measured with a DMM ? So the R_DC values ?

Will check as well on my unit.

Bye/thanks

galaxiex

Quote from: clintrubber on November 01, 2017, 09:10:23 AM
Quote from: clintrubber on November 01, 2017, 09:05:23 AM
Was wondering, any info on the reverb tank they used ? Input- & output-impedances ?

Oops, apologies, just saw the 225R & 222R indications in the last version of the schematic. Are these as measured with a DMM ? So the R_DC values ?

Will check as well on my unit.

Bye/thanks

Thanks, glad this is of some use.  :)

That's correct, DC resistance values measured with a DMM.

Yup, they are kinda cool little amps.  8) Actually sound pretty good for what they are.

Be aware there are at least 2 different versions of this amp,
both have the same name (National) and model number (GA920P) but different circuit boards/schematics.

I have one of each but not posted anything about the "other one" yet.
If it ain't broke I'll fix it until it is.

clintrubber

Quote from: galaxiex on November 01, 2017, 12:58:53 PM

Yup, they are kinda cool little amps.  8) Actually sound pretty good for what they are.

Just curious, how 'wet' can the reverb get at maximum?

QuoteBe aware there are at least 2 different versions of this amp,
both have the same name (National) and model number (GA920P) but different circuit boards/schematics.

I have one of each but not posted anything about the "other one" yet.

Hi,

Thanks, I'll check the R_DC of the spring unit here as well.

Mine looks alike (outside & inside) to your pics on page 1 at first I thought, but the placement is actually drastically different.
The sping tank is directly behind the front panel and the power devices are mounted on the base of the unit. The rear jack panel location is also different. Perhaps this is exactly describing your other unit ;-)

Mine has the blue section on the front as well, I've seen some pics of amps without.  I don't have the speaker enclosure.

FWIW/IIRC, I recall these amps being shown in mailorder catalogs. Here in The Netherlands the combo version went for 199 (Dutch guilders), and there was a 'Hondo II'  Les Paul copy for another 199 to be had, same catalog page, got us drooling :-)

OK, I'll see if it works, and if so, how it sounds. Curious to the spring reverb configuration. Too bad the enclosure doesn't permit a full sized tank.

Bye!

galaxiex

#24

QuoteJust curious, how 'wet' can the reverb get at maximum?


It can get pretty wet, but it doesn't sound good.  :(
I'm sure this is due to the limitation of the small tank.
It also doesn't help that there is only 1 spring.

I played with part values in mine and got a "fair" sounding reverb,
but it will never be a Fender Twin.

If you "over do it" with driving the tank too hard,
or increasing the gain on the recovery side, bad noises happen.
Screeching/feedback/metallic "clank" sound, etc.

Changing some cap values and possibly adding some caps in strategic places can smooth out the sound.
Limiting the frequency range that is fed to the tank helps. 

There are other small multi spring tanks around that would fit the chassis.
I never tried any, and there may be impedance matching problems with some of them.
No guarantees they will sound any better than the existing tank.

 
If it ain't broke I'll fix it until it is.

galaxiex

#25


QuoteMine looks alike (outside & inside) to your pics on page 1 at first I thought, but the placement is actually drastically different.
The sping tank is directly behind the front panel and the power devices are mounted on the base of the unit. The rear jack panel location is also different. Perhaps this is exactly describing your other unit ;-)


I don't recall the layout of the other amp.

Easy enough for me to open it up and get some pics.
I'll do that later.  ;)

Cheers!
If it ain't broke I'll fix it until it is.

clintrubber

Quote from: galaxiex on November 01, 2017, 06:51:23 PM
I don't recall the layout of the other amp.

Easy enough for me to open it up and get some pics.
I'll do that later.  ;)

Cheers!

Nice, who knows it's like mine :-)  And who knows that other one has the right schematic of your first one, and vice versa :-)     

I'll post some pics as well (this weekend)

Bye!

clintrubber

Quote from: galaxiex on November 01, 2017, 06:48:49 PM

QuoteJust curious, how 'wet' can the reverb get at maximum?


It can get pretty wet, but it doesn't sound good.  :(
I'm sure this is due to the limitation of the small tank.
It also doesn't help that there is only 1 spring.

OK, too bad. Only 1 spring, I see. So that would actually mean that a better small tank could still give some room for improvement...  but like you say, it'd be unrealistic to expect Twin Reverb sounds, let alone the superwet stuff as from the Fender Reverb unit.


QuoteI played with part values in mine and got a "fair" sounding reverb,
but it will never be a Fender Twin.

If you "over do it" with driving the tank too hard,
or increasing the gain on the recovery side, bad noises happen.
Screeching/feedback/metallic "clank" sound, etc.

Changing some cap values and possibly adding some caps in strategic places can smooth out the sound.
Limiting the frequency range that is fed to the tank helps. 

Thanks for sharing the results of your experiments, much appreciated!

QuoteThere are other small multi spring tanks around that would fit the chassis.
I never tried any, and there may be impedance matching problems with some of them.
No guarantees they will sound any better than the existing tank.

I have a broken Danelectro Spring King, but the tank itself is assumed to be OK.
That's a two spring medium length one, still not Fender-close, but probably a step up.

I realized that this 'Dano tank' matches the Fender TR & Reverb unit specs, w.r.t. the in- and output- impedances. Not the length obviously
Adding say the Surfy FET Reverb PCB (you'll have heard of it) would easily fit this enclosure.

Bye!

galaxiex

#28
Here's some pics of my "Blackface" version along with the schematic that matches this version.

I did change a few things... notably the tone control.
Sorry, I didn't document the changes but I used ideas from the AMZ notch tone control.

http://www.muzique.com/lab/notch.htm

Modding the tone control makes a significant difference to the tone of the amp.
For the better IMHO.

Having a little "scooped mid" helps for less muddy tone. Esp with distortion pedals.

I have seen the Surfy Bear reverb and if it could be incorporated into this amp
I think it may go a long way to improving the reverb tone.

Of the differences between this and my other "Silverface" version,
the input jack circuitry, reverb is transformer driven, and the different output devices.

As with all these amps, I added a 3 wire grounded cord and clipped the death cap.
If it ain't broke I'll fix it until it is.

clintrubber

Hi,

Thanks for the pictures! Interesting to see that 'they' (assuming it was all coming from the same factory) went through all these versions, colour schemes, component re-locations etc 

Mine is with the reverb tank located between the front and the PCB, so all pots are wired, so not directly on the PCB.
No reverb-TX.

Here's another, let's see what it does within a few minutes, now at $41 with 2 minutes to go.
https://www.ebay.nl/itm/263282000680?ul_noapp=true


Bye & I'll post pics from my GA920P this weekend