Grundig Amp for Spring Reverb

[ERRP55]

Hey Everyone,
This is my first post and my first project so please be patient with me haha. I am looking to learn as much as I can from you wizards.

I bought a Grundig Radio a few months ago and it peaked my interest because it came with a spring reverb amp/tank. Thankfully the owner had all of the original documentation so i also have a schematic for it which i will attach to this thread. I knew that i would most likely not use this spring reverb with the radio so I decided to pull it out in hopes that I could repurpose it as an outboard effect for recording/mixing studio use. However, I would like to make a couple modifications such as

- Installing an XLR input and XLR output.
- Implementing a balanced to unbalanced line input transformer at the front end of the amp.
- Installing an unbalanced to balanced line output transformer at the output.

My idea was to use a Cinemag CM-13102 for the input and output transformer since it is a 1:1 ratio. My only concern is how it would react with the 1Meg pot at the output of the amp. Maybe its best to remove the 1Meg pot? I would most likely feed the return of the reverb onto a separate channel on my mixer or separate track in my DAW so i could ride the volume via a fader.

I know i'll probably see very little gain in quality or benefits by installing these transformers, but its a project that i am hoping to learn from and apply concepts to future pieces of gear.

 

[metalockpick]

Cool.
Put the 1M reverb volume pot B4 the last ECC83 / 12AX7. Your output transformer is going to need a much higher impedance primary, the CM-DBX from Cinemag is a 12:1 that would work beautifully in that application.
I would consider using a rotary switch before c11 and having a selection of capacitors to act as a high-pass filter. I would consider putting a variable low pass filter (tone control) around C12/ r12. These would control how much bass and treble get into the springs.
I'm going to assume that you are going to take the output and send it into another channel. If so, you can use the channel EQ to dial the tone in.

 

[ERRP55]

Cool.
Put the 1M reverb volume pot B4 the last ECC83 / 12AX7. Your output transformer is going to need a much higher impedance primary, the CM-DBX from Cinemag is a 12:1 that would work beautifully in that application.
I would consider using a rotary switch before c11 and having a selection of capacitors to act as a high-pass filter. I would consider putting a variable low pass filter (tone control) around C12/ r12. These would control how much bass and treble get into the springs.
I'm going to assume that you are going to take the output and send it into another channel. If so, you can use the channel EQ to dial the tone in.

Thanks for the response. I like the idea of a hi-pass filter switch before c11 and the tone control around c12/r12 should be relatively simple. My only issue with the CM-DBX is that it will convert the unbalanced signal to mic level signal. Ideally I would want it to be line level, unless I misread the data on the CM-DBX.

 

[metalockpick]

Ah. the 4;1
CMLI10/600b would be the one to use.
You may have to change the output capacitors to a larger value to compensate for the impedance, then again, reverb isn't about the very low frequencies.
this will work

 

[ERRP55]

Ah. the 4;1
CMLI10/600b would be the one to use.
You may have to change the output capacitors to a larger value to compensate for the impedance, then again, reverb isn't about the very low frequencies.
this will work

I’ll check out the CMLI10/600b. How big of value increase would you change c19 and c18 to in order to compensate the impedance? I’m going to also sketch a schematic just to make sure I have everything correct.

 

[rock soderstrom]

I bought a Grundig Radio a few months ago and it peaked my interest because it came with a spring reverb amp/tank. Thankfully the owner had all of the original documentation so i also have a schematic for it which i will attach to this thread.

Wow, that's a cool device, probably also quite rare. What is the exact name of the radio and the reverb unit? (HV1+HS1?)

Is the whole thing functional as it is? I would test it first, watch out for the capacitors (kaboommm) when you switch it on, you might have to change a few!

The schematic says that the maximum input voltage is 100mV, which could be a minor problem in a modern studio environment. Maybe it makes sense to convert the first triode to cathode biasing ala Fender.

As I said, I would first test if you like the sound. A high-impedance input transformer can make sense, I would do the equalisation ITB or in the console, maybe convert the last triode to a cathode follower and do it without the output transformer. You do not need that much gain here, you can do this in the mixer or in your audiointerface, too. A slightly lower output impedance would be helpful, with or without OPT.

Nice photos would be very cool!

PS: What about background noise (humming and hissing), maybe you still have to do something there...

 

[ERRP55]

Wow, that's a cool device, probably also quite rare. What is the exact name of the radio and the reverb unit? (HV1+HS1?)

Is the whole thing functional as it is? I would test it first, watch out for the capacitors (kaboommm) when you switch it on, you might have to change a few!

The schematic says that the maximum input voltage is 100mV, which could be a minor problem in a modern studio environment. Maybe it makes sense to convert the first triode to cathode biasing ala Fender.

As I said, I would first test if you like the sound. A high-impedance input transformer can make sense, I would do the equalisation ITB or in the console, maybe convert the last triode to a cathode follower and do it without the output transformer. You do not need that much gain here, you can do this in the mixer or in your audiointerface, too. A slightly lower output impedance would be helpful, with or without OPT.

Nice photos would be very cool!

PS: What about background noise (humming and hissing), maybe you still have to do something there...

The amplifier came out of a Grundig SO 390, no exact name/model for the Reverb Amp. The Reverb Amp/Tank worked great with the radio and it was dead quiet not much hum or hiss. That’s about the only testing I did with it.

What benefits would I get with a cathode bias triode? With a cathode bias don’t I risk having less headroom and earlier distortion, which I guess is not completely bad, but for studio recording I would want it as clean as possible.

I’ll post pictures of the unit tomorrow morning. Thanks!

 

[rock soderstrom]

The amplifier came out of a Grundig SO 390, no exact name/model for the Reverb Amp.

Thanks!

What benefits would I get with a cathode bias triode? With a cathode bias don’t I risk having less headroom and earlier distortion, which I guess is not completely bad, but for studio recording I would want it as clean as possible.

I'm no expert, but I think it's the other way around. Cathode bias trumps grid leak bias in this respect. A Fender frontend can handle more than 100mV.

I found this about grid leak bias on the web:

"The reason this method is used is that it enables the cathode to be directly grounded. This reduced the effect of heater-cathode leakage in the days before DC heaters. It is used for low-level inputs. Large signals (or high source impedance) would create distortion, as the grid current is non-linear."

A test could show how serious the differences are in reality.

Just plug the reverb unit into the AUX path of your mixer, maybe it's not a problem at all.

 

[rock soderstrom]

I’ll post pictures of the unit tomorrow morning. Thanks!

Thanks. Here a youtube video with this beautiful Musiktruhe!



GnR

 

[rock soderstrom]

no exact name/model for the Reverb Amp

I think your reverb unit is the HV1 plus HS1.
HV1= Hall Verstärker 1; HS1 Hall System 1 (in englisch reverb amplifier and spring)
There was later a transitorised successor in a combo unit, the HVS1.

Here are some infos in german about the early tube version.

The following text states on the subject of max. input signal level:

"Only 20-50mV signal should reach the input of the reverb amplifier, only then is it guaranteed that the driver system is not overdriven."



These could also be purchased as free "Bausteine" (building blocks) to realise a customized system, follow this link:

http://www.hifimuseum.de/grundig-bausteine-1963-64.html


Phonomascope! Hail to the marketing department!

 

[ERRP55]

Thanks!


I'm no expert, but I think it's the other way around. Cathode bias trumps grid leak bias in this respect. A Fender frontend can handle more than 100mV.

I found this about grid leak bias on the web:

"The reason this method is used is that it enables the cathode to be directly grounded. This reduced the effect of heater-cathode leakage in the days before DC heaters. It is used for low-level inputs. Large signals (or high source impedance) would create distortion, as the grid current is non-linear."

A test could show how serious the differences are in reality.

Just plug the reverb unit into the AUX path of your mixer, maybe it's not a problem at all.

Awesome I’ll definitely add the cathode bias resistor/cap to the triode of the ECL86. My plan today is to draft up a schematic so I will attach that later this afternoon.

 

[ERRP55]

I think your reverb unit is the HV1 plus HS1.
HV1= Hall Verstärker 1; HS1 Hall System 1 (in englisch reverb amplifier and spring)
There was later a transitorised successor in a combo unit, the HVS1.

Here are some infos in german about the early tube version.

The following text states on the subject of max. input signal level:

"Only 20-50mV signal should reach the input of the reverb amplifier, only then is it guaranteed that the driver system is not overdriven."

View attachment 110528

These could also be purchased as free "Bausteine" (building blocks) to realise a customized system, follow this link:

http://www.hifimuseum.de/grundig-bausteine-1963-64.html
View attachment 110529

Phonomascope! Hail to the marketing department!

A line level signal is roughly 300-1000mV correct? A +4dbu signal is about 1V. if my max clean signal for this amp is 50mV, will a correctly calculated bias resistor/cap on the triode be enough to get me to close to 300mV or past that? Thanks!

 

[Chandra]

If the original unit was fitted inside together with the speakers, the vibrations from the (speaker) cabinet would have some added effect, though the reverb is mounted on vibration arrestors.. Just guessing from my hearing of guitar amplifier cabinets.
Tell us how you feel the effect.
Regards.

 

[rock soderstrom]

A line level signal is roughly 300-1000mV correct? A +4dbu signal is about 1V. if my max clean signal for this amp is 50mV, will a correctly calculated bias resistor/cap on the triode be enough to get me to close to 300mV or past that? Thanks!

Good question, perhaps one of the experts can say something about this.

What I can contribute is this. Years ago I built this project, which is very similar to your reverb unit. Ignore the PCL86, it's just an ECL86 with a different heater, otherwise completely identical.



The VTR63 worked very well and was somewhat similar to the Fender Spring Reverb Tank 6G15. I used it mainly for guitar, but also had it looped into my AUX path several times, which worked. How much level this circuit could really accept undistorted, I can't tell you today. In principle, you can adjust the signal with a voltage divider or potentiometer. The goal should still be to not have to attenuate the signal too much, otherwise the SNR suffers.

For those interested, I still have the original article, unfortunately only in German.

Edit:

I'm not sure if you need a mix function, I would leave it out and hardwire the cathode follower behind the recovery amplifier.

 

[nsalis]

The reason why the HV-1 amplifier has a high input sensitivity (20 ... 50 mV) is that you can put the voltage divider (shown in the supplement to "rock soderstrom") in front of the amplifier.

This voltage divider has a very high-impedance input (left channel and right channel of the source connected with 1M ohms each, then shunt resistor of 50k ohms to ground). A voltage divider with such a high-impedance input was necessary so that the crosstalk between the radio’s stereo channels was not reduced. The disadvantage of the voltage divider is that it reduces the level considerably, namely by 26 dB.
Or to put it another way: The voltage emitted by the radio and sent to the voltage divider was 26 dB higher. If you take the mean value of the specified input voltage range (20 ... 50 mV), you get 35 mV. So the level emitted by the radio was 26 dB higher than approx. 35 mV. This corresponds to a voltage of 700 mV.
If you want to feed a studio line-level signal of 4 dBu = 1.23 V into the Grundig HV-1 without it starting to distort, you have to attenuate it by about 30 dB. Such an attenuation can be easily achieved with a product like the Radial ProAV1 Direct Box. It has a balanced XLR input. On the XLR output connect pin 3 (cold) and pin 1 (cable shield) and connect to Grundig’s 3-pin-DIN connector pin 2 (ground). XLR pin 2 (hot) connects to Grundig’s 3-pin DIN connector pin 1 (audio input).

As far as the output of the HV-1 is concerned, there is a maximum output voltage of approx. 13 V here. If you subtract the usual headroom of 14 dB, you can expect an undistorted nominal voltage of approx. 2.5 volts. That would more or less correlate with the 4 dBu studio level, but this voltage only arises if you feed a destination with high impedance (more than 100k ohms) input. This is not the case with modern studio equipment. So you would have to use a buffer amplifier here that has a relatively high impedance input.

A good product would be the Sonifex RB-Li2. It has 100k ohms input impedance and a transformer isolated line level output. The gain is adjustable. Connect Grundig’s DIN connector pin 2 (Gnd) to cable shield and XLR input pin 1 and pin 3 of the Sonifex. Connect Grundig’s DIN connector pin 3 (audio output) to Sonifex’s XLR input pin 2.

https://www.sonifex.co.uk/redbox/rbli2_ld.shtml
ns

 

[ERRP55]

The reason why the HV-1 amplifier has a high input sensitivity (20 ... 50 mV) is that you can put the voltage divider (shown in the supplement to "rock soderstrom") in front of the amplifier.

This voltage divider has a very high-impedance input (left channel and right channel of the source connected with 1M ohms each, then shunt resistor of 50k ohms to ground). A voltage divider with such a high-impedance input was necessary so that the crosstalk between the radio’s stereo channels was not reduced. The disadvantage of the voltage divider is that it reduces the level considerably, namely by 26 dB.
Or to put it another way: The voltage emitted by the radio and sent to the voltage divider was 26 dB higher. If you take the mean value of the specified input voltage range (20 ... 50 mV), you get 35 mV. So the level emitted by the radio was 26 dB higher than approx. 35 mV. This corresponds to a voltage of 700 mV.
If you want to feed a studio line-level signal of 4 dBu = 1.23 V into the Grundig HV-1 without it starting to distort, you have to attenuate it by about 30 dB. Such an attenuation can be easily achieved with a product like the Radial ProAV1 Direct Box. It has a balanced XLR input. On the XLR output connect pin 3 (cold) and pin 1 (cable shield) and connect to Grundig’s 3-pin-DIN connector pin 2 (ground). XLR pin 2 (hot) connects to Grundig’s 3-pin DIN connector pin 1 (audio input).

As far as the output of the HV-1 is concerned, there is a maximum output voltage of approx. 13 V here. If you subtract the usual headroom of 14 dB, you can expect an undistorted nominal voltage of approx. 2.5 volts. That would more or less correlate with the 4 dBu studio level, but this voltage only arises if you feed a destination with high impedance (more than 100k ohms) input. This is not the case with modern studio equipment. So you would have to use a buffer amplifier here that has a relatively high impedance input.

A good product would be the Sonifex RB-Li2. It has 100k ohms input impedance and a transformer isolated line level output. The gain is adjustable. Connect Grundig’s DIN connector pin 2 (Gnd) to cable shield and XLR input pin 1 and pin 3 of the Sonifex. Connect Grundig’s DIN connector pin 3 (audio output) to Sonifex’s XLR input pin 2.

https://www.sonifex.co.uk/redbox/rbli2_ld.shtml
ns

nsalis thanks for sharing your wisdom and explanation! I do have a couple questions still. Regarding the input voltage, if I am understanding you correctly. Changing/adding a voltage divider to the input that will attenuate the signal by 30db is the only way to feed a proper line level signal to the of the ECL86 triode without the need for a transformer? Or should i disregard the 100mV rating because this only applies to the voltage divider that is present when connected to the radio.

My thoughts were to add a CMLI-10/600 to convert the 600ohm line input signal to -10 level source that feeds the ECL86 triode. Will this not work and cause SNR issues? I was also thinking of using the CM-13102 at the output since it is a high impedance to line transformer. I understand that I could use the ProAV1 DI and the Sonifex, but the purpose for this project is to rehouse this unit into a 1U rack chassis to use it as a studio outboard effect. This is my first major mod/diy project so I would really like to have the experience of building this unit from scratch.

 

[metalockpick]

Running the input xfrmr as a 4:1 stepdown (10k side input, 600 Ohm side to the first tube) will work to reduce the level by 12 dB, giving you isolation and a balanced input. You shouldn't have any noise issues.

 

[ERRP55]

Running the input xfrmr as a 4:1 stepdown (10k side input, 600 Ohm side to the first tube) will work to reduce the level by 12 dB, giving you isolation and a balanced input. You shouldn't have any noise issues.

Shouldn't I reverse the transformer? 600ohm input side, 10k side to first tube? If I am going to use this reverb as an insert on a mixer, the signal that I am feeding it is line level. So, won't it want to see the 600ohm input side of the xfrmr that way the HV-1 reverb see's the unbalanced 10k high impedance?

 

[Brian Roth]

If you "reverse" the transformer as you describe, then it will provide a step UP in voltage, which you don't want/need.

Bri

 

[ERRP55]

If you "reverse" the transformer as you describe, then it will provide a step UP in voltage, which you don't want/need.

Bri

Stupid me I was looking at it complete backwards. Thanks!