John Linsley Hood preamp designs
A versatile and, thankfully, prolific designer John Linsley Hood has educated many with a fresh eye, backed by a much-welcomed and thorough empiricism. Unfortunately, some souces can contain errors and where possible these will be highlighted.
A zip-file of data on JLHs' Modular Pre-Amplifier Design (Wireless World article, July 1969), contemporary to the simple class A power amp design published in April of that year (circuit diagram) is available on request.
Below is Linsley Hoods' 'Liniac' RIAA stage (WW, Sept '71, data kindly supplied by Cyril Wilkin).
This configuration explored the possibilities utilising low-level Darlington transistors to create inverting high input impedance, high gain, low noise and low distortion building blocks. Other 'flat' approaches included
and a low distortion oscillator.
In the same article, attention was paid to other configurations.
The 75W Powertran kit used a 'Liniac' preamp with a 'mode' switch and switched range tone control with variable slope high filter and switched low filter (PDF available). A later 'Deluxe' version made construction easier with plug-in boards, the power amp used being similar to the improved HiFi News and Record Review (1972?) version and incorporated other improvements like a headphone socket. Heatsinking was limited and some had problems with assembly which could result in accidental gross shorts. Although flexible, with in excess of 25 switch contacts per channel mechanical noise could be problematic, as could mains borne interference despite a steel enclosure and screened mains transformer, notwithstanding the optimistic output rating which had not been intended as a continuous one. If these designs (or others) are being built today it is recommended that the DIN connector convention be abandoned, RCA phono connectors being substituted instead.
Kits like these and the Curtis (Jelgate) ones could require a commitment that was not envisaged by the designers because of the intervening kit suppliers' designers limitations. For the uninitiated or unqualified (or qualified!) enigmas could arise when the expectation was of perfection which was not matched by reality. For an experienced hardware engineer these could prove challenging on occasion but ultimately gave experience beyond expectation and a lift when the 'box' finally sang.
Amongst other designs, this class A headphone amplifier appeared in 1979.
Because of the high quiescent current JLH recommended T03 style regulators (MC7812 CK, MC7912 CK) regulators, smoothed by 1mF and decoupled by 47µF per rail.
The JLH 'ripple eater' is covered adequately by Jon Clancy. This can offer the equivalent of 0.5F smoothing.
Following the Wireless World '80-100W Mosfet Audio Amplifier' (Aug '82), which many found interesting despite some very serious errors (corrected in the September '82 issue, p63), the 1982/3 Modular Preamplifier design (another excellent source of food for thought) appeared whose schematics are covered here together with some of the later ones (ETI, etc) marketed as Hart kits. Note that it is recommended that both sides of any coupling capacitors be grounded via high value resistors to prevent switching 'clicks'.
The later design was powered by a conventional 7815/7915 regulated supply. The main intention was that the modules, switched into the signal path as required, would function at a nominal 0.774V, 600R impedance. A virtual earth mixer was used to select the inputs and provide a balance function, the gain pots being ganged. JLH expressed regret at the high impedances used, however, with the DIN convention these are unavoidable.
Some designs seen used two sets to accommodate DIN and RCA phono conventions.
This arrangement for a moving-coil buffer has been used with success as a mic preamp and is one of the simplest MC buffer designs seen. Electrolytics can be paralleled with 1µF polycarbs. Switching contacts, if used, should be gold-on-gold. The matching RIAA stage is given below. Remove 1µF output cap if used with this.
Employing a 'series-shunt' two-stage configuration, this approach allows accurate tracking of the RIAA curve with standard value components and an improved upper frequency response, without being over-engineered.
The noise figure of 450 ohms does not refer to the layout shown and if difficulties are encountered sourcing the BC447/BC448 transistors, their JEDEC equivalents are 2N2907A/2N5551.
A combination of one or more of the above is considered adequate for all practical purposes, barring the possible substitution of the opamp. But then there's the more complex discrete version below, if a greater degree of precision is deemed necessary. Some other items of interest from the '82/3 Modular Preamplifier designs;
a microphone preamp,
an image-width module (other designs),
an impulse-noise blanker (useful for scratched vinyl),
low (scratch) and high (rumble) pass bootstrap filters for a third order flat response,
a 'Clapham Junction' or spot-frequency step lift and cut-type tone control,
and a class A headphone amplifier.
Simple level meters were included and, as mentioned above, the matching power amp would have been the 80-100W mosfet described in Wireless World, August '82.
Basic feedback arrangement for a RIAA design giving a close match. Personal preference would include input coupling and feedback decoupling caps to give -3dB roll-offs at about 15Hz or more.
The more complex discrete RIAA stage (Hart JLH K1450, 1995, PDF of construction manual available) intended for the K1400 preamp.
Switching contacts, if used, should be gold-on-gold. Ideally matched to a cartridge in its' deck, thus negating use of contacts. The 'isolated' input is intended for moving magnets and the 'direct' one for moving coils. As with other designs some data may contain minor errors (eg; mc sensitivity) but was that actually supplied. It was an idea to check all available sources before construction. Allow 30mA per rail. If a differential cable run is required, then an IC solution can be argued to perform better than transformers, eg: SSM2142/3 (s/n >105dB, 7MHz). A 100mA supply for a complete stereo solution should suffice.
Vincent Hawtin notes that "... the 2n2 and 680pf amount to 2880pf this time constant should be with the 1meg resistor 2937us thus adding an extra 50 pf to the 2880pf will give 2930us about as close as you can get with standard values John always tried to get as close with easily available parts so 2880 would be near enough but there is a slight improvement with the added 50 pf which reduces the slight mid lift without it the rest of the arrangement is as good as it can get without special values. One other change that was in the K1400 kit was the 12 ohm with the moving coil switch on the I/P amp was increased to 25 ohms to improve the low frequency roll off in the MC position to 14 hz with the switch in the MM position this would be 2.2hz as you can see with 470uf and 12 ohms not so much bass. You can X-check this mod in his book Audio Electronics.. page 126 refers to diagram 1f in preceeding page of networks thus :- 1meg /82k =12.195.......should be 12.38 near enough 82k and 1n = 82us..... should be 81.1us near enough and finally the 2.2+680 with 1 meg =2880us... should be 2937us so adding the 50 pf will do it nicely... ... yes you are quite correct that the values on my cct specify 2.5% tolerance for caps and 1% for resistors and that would certainly take care of the 50 PF extra but when i measured the caps from Hart they were pretty much spot on the values so i presume they went one better and used 1%.. My own design of RIAA using op amps uses 1% tolerance caps and the design is very accurate in the RIAA curve and so i use it to check other manufacturers pre-amps some of which do not sound quite right I believe that with very good amplification smaller errors are noticable.. By the way in the early gain stage the 470 uf caps sent with the Hart kit were NOVO types when these are replaced by the low Z Nichicons this amp takes on a new persona.. Also if you can place a cap of about 0.22 to 0.33uf straight across the 5volt rails + to - that also helps to clean the signal seems to cut some noise from those 5volt regs.."
For the 50pF value, 3 //'d 18pF Philips 683 series (ceramic plate, low K, 2%, 100V, 0ppm/°C) could be used. Regulator noise (typ. 40µV, ripple rejection 60dB for L series) can be overcome by using simple zeners or additional decoupling, however, a simple Armstrong solution can be effective. The later K1500 below uses dual IC opamps reducing the component count and complexity considerably and is recommended for this approach.
Some circuits from the K1400 preamp (PDF of construction manual available); a low back-EMF relay driver,
main buffer (one version has a 1k pot replacing the 150Rs, the wiper going to the o/p C to cope with device differences)
switched range Baxandall tone controls (opamps specified for these, the tape buffer and filter were TL072, LF353, LM833, OP27/37),
rumble filter (for calculations see above)
and the headphone amplifier.
These and the Hart K2100 'Chiara' single ended class A headphone amplifier (below, PDF of construction manual available) could be run from a single regulated supply, the K3500 'Andante' (PDF of construction manual available), via 9/25way D connectors. This comprised of two sets of 15V 78/9 series regulators (separate pairs for the preamp and headphone amp) with a separate 7815 for switching relays, smoothed by 2m2F, decoupled by 100µF. As with a number of kits (Hart and Powertran), mains filtering was poor.
The 'separates' designs had PCB selectable ground options. If a ground is lifted, an HF path to chassis (10nF ceramic capacitor) usually improves and is best made permanent.
The Hart K1500 RIAA (Aug, '91), a good source of information was found here.
This design seemed to acknowledge that opamp technology had subjectively caught up with discrete designs although some considered that i/p AC coupling could improve (LF roll-off to cater for deck deficiencies, first stage roll-off 18Hz for MC and 16Hz for MM, and 13Hz for the second stage suggesting a value of 0.18 and 0.27µF between the i/p loads). The schematic was originally published in the November 1990 edition of "Gramophone". Note the discussion of the values of identical second stage components above (K1450). Power was supplied via an inadequate 5-pin DIN connector and positions on the PCB were available for 100mA regulator ICs (78L12 and 79L12). 1M Rs across the switches could reduce switching noise. The later LM4562, or similar, would be a good choice for the opamps.
Other interesting JLH circuits include a millivoltmeter,
a moving-coil preamp and a symmetrical, low-distortion, gain block (circuit information for both courtesy of Michael Jonassen) detailed in 'Symmetry in audio amplifier circuitry', E&WW, Jan '85, pp31-34.
JLH simple class A, MOSFET and other RIAA designs
Google search for John Linsley Hood.
Contact me at firstname.lastname@example.org
especially if you want additional content to this page
or if you find any links that don't work. Don't forget
to add the page title or URL. Take care!
Back to index, sound, tips or home.