Grampian Spring reverb

Just found the Grampian Spring schem ,

I was aware Pete Townsend had one of these in his rig ,but never realised what it was .
Turns out to have a very neat light bulb limiter , theres a demo towards the end of  this vid .

https://www.youtube.com/watch?v=s0gRvSNx2g4

The effect on the snare is very awesome indeed , Im defo gonna need some of that hot sauce in my home brew tube reverb .

The mic preamp is interesting too. First time I saw it was in early .Audio Development mixers. I don't know who originated it but it seems to be a British design, that I haven't seen anywhere else. The xfmr secondary is part of the NFB loop, so gain and noise are variable in accord to the mic impedance; particularly, noise is minimum when the input is disconnected.

Hi Abbey,
There is a hammond reverb that seems to utilise feedback around  the spring too  ,  its difficult for me to visualise how it effects things inside the loop or whats exactly is going on . The grampian seems kind of unique in that it includes the bulb in the feedback path .

From the demo video I posted the 'Gramp' seems to sound the best of the bunch, probably the only one with a long tank also .Scratch Perry used it to good effect on many of his recordings , including punching the unit for the signature metalic 'clanging' sound.

Of course the bulb limiting is rough as the badgers proverbials ,but the artifacts it leaves behind do add lots of colour to the reverb sound .

I want to make the tanks on my reverb driveable from any speaker output ,so I'll put an attenuator that encompasses the bulb limiter and make that switchable , a bridged T comprising  dual 200 ohm pots gives me a range of about 20db for an 8 ohm load, I'll just arrange so any surplus power is dumpded before the bulb ,now I can effctively enough change the threshold of the compression . 

The other post got me thinking  ,could an ldr driven by the  lightbulb  and say controling feedback to a previous stage form a simple attack control ?

Tubetec said:

Hi Abbey,
There is a hammond reverb that seems to utilise feedback around  the spring too  ,  its difficult for me to visualise how it effects things inside the loop or whats exactly is going on .

Click to expand...

A good example of current-to-current NFB. Current is sensed via the 1.8 ohm resistor, thus converted as a voltage, that is fedback as a current via the 3.3k to the base of Q6, against the drive current from Q5.

  The grampian seems kind of unique in that it includes the bulb in the feedback path .

Click to expand...

Since it's teh current that is sensed, whatever resistance in teh circuit comes unnoticed (within limits)

  From the demo video I posted the 'Gramp' seems to sound the best of the bunch,

Click to expand...

Yes, it's quite obvious.

probably the only one with a long tank also .

Click to expand...

Are they all spring-based? It seemed to me the Pioneer SR202 sounded more like a delay. There was a number of reverb/echo units that used an acoustic delay line consisting of a folded tube with a speaker and a mic.

Of course the bulb limiting is rough as the badgers proverbials ,but the artifacts it leaves behind do add lots of colour to the reverb sound .

Click to expand...

Actually I was surprised how good it sounds when teh bulb illuminates, which indicates a serious drive power.

The other post got me thinking  ,could an ldr driven by the  lightbulb  and say controling feedback to a previous stage form a simple attack control ?

Click to expand...

That would be double compression, from the bulb itself increasing its resistance and from the gain reduction due to the ldr.
You could try it; put the LDR at the input of the "buffer/splitter" stage.
I'm not sure it would result in a proper attack control, but it would certainly increase compression.

I think they were all spring based units in that demo .

Maybe some sort of crude expansion circuit tied into the recovery stage might be more appropriate and interesting sounding .

Tubetec said:

There is a hammond reverb that seems to utilise feedback around  the spring too  ,  its difficult for me to visualise how it effects things inside the loop or whats exactly is going on .

Click to expand...

Well, the Accutronics site writes about current drive being the ultimate. I have tried both current and voltage drive. Indeed, due to the inductive nature of the transducers, HF shelving must be applied to voltage drive in order to achieve the desired response, but when this is done, I can't hear any difference in the recovered signal. Actually, the only difference is that voltage drives increases the damping of the drive transducer, but it is so low to start with that it makes no audible difference.

I have various tanks with different input/output transducers , for simplicity I want to use 8 ohm drive coils , this also means my driver amp can double duty as small speaker or headphone  amp where the need arrises .

Ive tried higher impedance drive coils off op amp circuits but I found with valve drive and subsequent attenuation/limiting  before the transducer, degrees of dirt could be dialed in to the drive stage , the drive signal then attenuated so the signal doesnt overload the input transducer , I could get a whole range of different colours to the reverb this way .

Multiple tanks with summed outputs also noticeably smoothed out metalic clang/echo ,also resulted in more even, natural and complex  decay  . . A balanced recovery stage with 600 ohm impedence coils worked out good into standard mic inputs ,typically way better than  hi z valve recovery ,but standard guitar amps are a very noisey environment to run a spring with all kinds of induced hums from nearby mains transformers .ac heaters etc. Hopefully a seperate powersupply enclosure remote from the springs and amps will allow much better noise performance both on the drive and recovery end.

I see in the hammond unit with feedback across the spring itself a filter network changes the damping and level the spring is driven at for different reverb  tone settings, not many made use of this in tank reverb design .

I was thinking  about a regen control that feeds some spring sound output back to the drive amp , Ive tried this a few times by driving an attenuator and spring directly  across  the speaker of a combo  amp ,then the spring output brought back into a second input channel on the amp ,I noticed I could make very long reverbs with carefull setting of the controls, it also tended to further difuse the boingy ping pong charachter we associate with mechanical springs ,the  tails seemed to have finer resolution/ less grittiness also. The problem with regen is it can easily break into oscillation ,maybe an LDR in the side chain might be interesting .

OMG That germanium smashing
I'm in love. I need this in my life !

The distortion and bulb does give quite a crack to even a thin synthetic snare ,
The bulb effectively prevents the drive transducer of the tank from overloading , a good thing as it doesnt sound much good when it does . You can dial in as much distortion as you like to add an extra dimension or sparkle to the sound .

Rod Elliots site has some info on a simple lamp driven tank circuit here ,
http://sound.whsites.net/articles/reverb.htm
Im not sure why Rob didnt put the bulb in series with the tank ,other than that it does bear a striking resemblance to the Grampian drive setup .

Ive have a few tanks set up lately for test purposes , So from a small audio mixer I connected the main outputs to a pair of Accutronics  250 ohm input 2250 ohms output tanks , I had initially tried mic in as return but the low loading of the output  transducer didnt allow the spring to sing out properly . Ive now used a pair of active DI boxes to present a couple of meg ohms to the output transducers of the tanks , tons more level and sparkle although the basic gain of the Di box at 20 db is still a little on the low side . The springs being run off the main outputs and returned on a channel theres always some feedback but the loop lenghtens the reverb time  and smoothes the edges in the decays ,theres a delay of about 30-50 ms across the spring ,a small proportion of the output signal fed back adds a  thickening the effect .

At the moment I just have the drive  transducers connected unballanced to my mixer outputs but if I was to untie one side of the input transducer from ground and provide balanced connection  could I expect an increase in drive levels ?
I have some 6.3 volt lamps , the usual kind you find as backlighting or indicators on old gear , they work well as limiter on  the speaker line of smaller powered amps ,but I never tried them on line level or op amp type drive impedences .

Hi Das ,
glad you enjoyed the sound ,  more like a gunshot in a cathederal than a snare ,

Often little stand alone record players from the mid to late 60's have Germainium amps  built in , three transistors and a driver transformer , exactly whats in the Grampian . I have a small board from one of these in the shed with mullard GET transistors , I might dig that out for fun .
Really once you have yourself a tank to experiment with you can easily make some kind of a noise . The average hifi reverb ad-on in the video clip sounded bad , the Grampian was the only one of the bunch that provided  overdrive and compression on the springs inputs . I think many of the guitarists who used the grampian also used it as a gain booster/distortion on their direct signal as well as reverb .

Even though in the modular synth community spring reverbs are very popular ,very few if any of the modern offerings do much sonically for me , some intentionally overload the input transducer and call that a sound , most have feedback control which is nice , but they all seem to have completely missed unleashing the real animal in the tank , by making a decent spring driver stage with loads of power/ nice overload and placing a lamp and attenuator on the line so you can prevent satuaration of the coil/core . or add compression .

A little custom pcb/eurorack  front panel  could easily give Doepfers A-199 a good sonic punch in the face  at 100 bucks , Discrete transistor drive  off an op amp like in Rods setup , the lamp ,the attenuator ,a cv controllable filter on the return .A reverb kit for the modular and home constructor would sit up very nicely in the thousands of other synth gadgets .

I just had a chance to try a brand new Belton short tank , two spring model , compared to the old Accutronics it has a good bit more output, coil impedances more or less the same , the Belton core and coil assemblies look great everythings neat and tight and the coil wires attach in direct to the winding ,not a plastic plug like later Accu's. The small tank Belton is as good as or better sounding than any of the long tank accutronics units I have, ,definately no quality control problems , and the core/coil  is screwed together  not rivited like the Accu's ,so makes it easier to rewind if you need to for a different impedance or because the coil is  open.

Tubetec said:

Im not sure why Rob didnt put the bulb in series with the tank ,

Click to expand...

Actually the bulb is in series with the tank, since the negative input of the opamp is very high impedance.

Ahhh ,so it is in series , 
I just wasnt looking at it right  ;D
maybe its the way its drawn threw me .

The grampian style circuit might have the edge in terms of niceness of overload ,with the interstage transformer and fizzly GET's ,but Rods circuit is a perfect modern equivalent using an op amp in place of iron , faithfull to both the original Hammond principle (or whoever he pinched it off ) and how its done in the Grampian . Looks like every spring maker for the last 50 years didnt have that 'lightbulb' moment ,and settled for feeble drive off an op amp at higher impedance 

I have a Grampian, but don't remember it sounding that great. I will have to pull it out and try it again. As I recall when the light bulb goes off it draws enough power to fuzz out the output, which I figured it what made it nice for guitar as a level triggered fuzz with mild reverb.

Or perhaps dried out caps did not have enough reserves to cover the increased draw?

mjrippe said:

Or perhaps dried out caps did not have enough reserves to cover the increased draw?

Click to expand...

When the bulb heats, its resistance increase, but since the drive amp is a current driver, it increases the voltage until it hits the rails, so even with perfectly healthy components, it is prone to distort.

Thanks Abbey, that makes sense.